diff --git "a/checkpoint-1260/trainer_state.json" "b/checkpoint-1260/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1260/trainer_state.json"
@@ -0,0 +1,8958 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9940546967895365,
+  "eval_steps": 105,
+  "global_step": 1260,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0023781212841854932,
+      "grad_norm": 32.74397118558861,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.9478,
+      "step": 1
+    },
+    {
+      "epoch": 0.0023781212841854932,
+      "eval_loss": 3.373392343521118,
+      "eval_runtime": 78.9756,
+      "eval_samples_per_second": 8.091,
+      "eval_steps_per_second": 1.013,
+      "step": 1
+    },
+    {
+      "epoch": 0.0047562425683709865,
+      "grad_norm": 37.080911592721954,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.8264,
+      "step": 2
+    },
+    {
+      "epoch": 0.007134363852556481,
+      "grad_norm": 31.107267997702266,
+      "learning_rate": 1.5e-06,
+      "loss": 2.9936,
+      "step": 3
+    },
+    {
+      "epoch": 0.009512485136741973,
+      "grad_norm": 27.574905774161167,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.8944,
+      "step": 4
+    },
+    {
+      "epoch": 0.011890606420927468,
+      "grad_norm": 13.643957484299273,
+      "learning_rate": 2.5e-06,
+      "loss": 2.7687,
+      "step": 5
+    },
+    {
+      "epoch": 0.014268727705112961,
+      "grad_norm": 15.077028980653411,
+      "learning_rate": 3e-06,
+      "loss": 2.6623,
+      "step": 6
+    },
+    {
+      "epoch": 0.016646848989298454,
+      "grad_norm": 14.569557474559408,
+      "learning_rate": 3.5e-06,
+      "loss": 2.9007,
+      "step": 7
+    },
+    {
+      "epoch": 0.019024970273483946,
+      "grad_norm": 13.894984550517007,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.8359,
+      "step": 8
+    },
+    {
+      "epoch": 0.02140309155766944,
+      "grad_norm": 13.546442233583257,
+      "learning_rate": 4.5e-06,
+      "loss": 2.7583,
+      "step": 9
+    },
+    {
+      "epoch": 0.023781212841854936,
+      "grad_norm": 11.166476039936938,
+      "learning_rate": 5e-06,
+      "loss": 2.518,
+      "step": 10
+    },
+    {
+      "epoch": 0.026159334126040427,
+      "grad_norm": 7.787448179397784,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 2.6494,
+      "step": 11
+    },
+    {
+      "epoch": 0.028537455410225922,
+      "grad_norm": 4.72349822440695,
+      "learning_rate": 6e-06,
+      "loss": 2.4022,
+      "step": 12
+    },
+    {
+      "epoch": 0.030915576694411414,
+      "grad_norm": 4.100722460414476,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 2.3933,
+      "step": 13
+    },
+    {
+      "epoch": 0.03329369797859691,
+      "grad_norm": 4.193151112965372,
+      "learning_rate": 7e-06,
+      "loss": 2.5468,
+      "step": 14
+    },
+    {
+      "epoch": 0.0356718192627824,
+      "grad_norm": 5.502246954578136,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 2.5126,
+      "step": 15
+    },
+    {
+      "epoch": 0.03804994054696789,
+      "grad_norm": 5.716937946349337,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.6761,
+      "step": 16
+    },
+    {
+      "epoch": 0.04042806183115339,
+      "grad_norm": 3.008867017303434,
+      "learning_rate": 8.5e-06,
+      "loss": 2.4264,
+      "step": 17
+    },
+    {
+      "epoch": 0.04280618311533888,
+      "grad_norm": 3.6016120293217178,
+      "learning_rate": 9e-06,
+      "loss": 2.3836,
+      "step": 18
+    },
+    {
+      "epoch": 0.04518430439952437,
+      "grad_norm": 2.1431386543975908,
+      "learning_rate": 9.5e-06,
+      "loss": 2.4879,
+      "step": 19
+    },
+    {
+      "epoch": 0.04756242568370987,
+      "grad_norm": 3.838539096237921,
+      "learning_rate": 1e-05,
+      "loss": 2.4116,
+      "step": 20
+    },
+    {
+      "epoch": 0.04994054696789536,
+      "grad_norm": 1.9119200890173822,
+      "learning_rate": 1.0500000000000001e-05,
+      "loss": 2.6716,
+      "step": 21
+    },
+    {
+      "epoch": 0.052318668252080855,
+      "grad_norm": 2.0853737807318904,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 2.4523,
+      "step": 22
+    },
+    {
+      "epoch": 0.054696789536266346,
+      "grad_norm": 1.6109539551135528,
+      "learning_rate": 1.15e-05,
+      "loss": 2.3086,
+      "step": 23
+    },
+    {
+      "epoch": 0.057074910820451845,
+      "grad_norm": 2.575933824126331,
+      "learning_rate": 1.2e-05,
+      "loss": 2.5576,
+      "step": 24
+    },
+    {
+      "epoch": 0.059453032104637336,
+      "grad_norm": 1.4827660467902501,
+      "learning_rate": 1.25e-05,
+      "loss": 2.264,
+      "step": 25
+    },
+    {
+      "epoch": 0.06183115338882283,
+      "grad_norm": 2.1022593224903128,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 2.8781,
+      "step": 26
+    },
+    {
+      "epoch": 0.06420927467300833,
+      "grad_norm": 2.756765764204405,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 2.5071,
+      "step": 27
+    },
+    {
+      "epoch": 0.06658739595719382,
+      "grad_norm": 1.6249455774563617,
+      "learning_rate": 1.4e-05,
+      "loss": 2.5085,
+      "step": 28
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 1.8653671914008603,
+      "learning_rate": 1.45e-05,
+      "loss": 2.4368,
+      "step": 29
+    },
+    {
+      "epoch": 0.0713436385255648,
+      "grad_norm": 1.867632087572305,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 2.1204,
+      "step": 30
+    },
+    {
+      "epoch": 0.07372175980975029,
+      "grad_norm": 1.2367329590346625,
+      "learning_rate": 1.55e-05,
+      "loss": 2.1118,
+      "step": 31
+    },
+    {
+      "epoch": 0.07609988109393578,
+      "grad_norm": 1.5969207000997636,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.4747,
+      "step": 32
+    },
+    {
+      "epoch": 0.07847800237812129,
+      "grad_norm": 1.5075133034220278,
+      "learning_rate": 1.65e-05,
+      "loss": 2.3349,
+      "step": 33
+    },
+    {
+      "epoch": 0.08085612366230678,
+      "grad_norm": 1.4803692479229955,
+      "learning_rate": 1.7e-05,
+      "loss": 2.3915,
+      "step": 34
+    },
+    {
+      "epoch": 0.08323424494649227,
+      "grad_norm": 2.692949771759104,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 2.6397,
+      "step": 35
+    },
+    {
+      "epoch": 0.08561236623067776,
+      "grad_norm": 1.3636512770329847,
+      "learning_rate": 1.8e-05,
+      "loss": 2.2944,
+      "step": 36
+    },
+    {
+      "epoch": 0.08799048751486326,
+      "grad_norm": 1.8310086696195464,
+      "learning_rate": 1.8500000000000002e-05,
+      "loss": 2.2614,
+      "step": 37
+    },
+    {
+      "epoch": 0.09036860879904875,
+      "grad_norm": 1.7765027708264853,
+      "learning_rate": 1.9e-05,
+      "loss": 2.3579,
+      "step": 38
+    },
+    {
+      "epoch": 0.09274673008323424,
+      "grad_norm": 1.4484769960901491,
+      "learning_rate": 1.95e-05,
+      "loss": 2.4548,
+      "step": 39
+    },
+    {
+      "epoch": 0.09512485136741974,
+      "grad_norm": 3.1520205275209414,
+      "learning_rate": 2e-05,
+      "loss": 2.5208,
+      "step": 40
+    },
+    {
+      "epoch": 0.09750297265160524,
+      "grad_norm": 1.5897739849482102,
+      "learning_rate": 1.9999981652287733e-05,
+      "loss": 2.4216,
+      "step": 41
+    },
+    {
+      "epoch": 0.09988109393579073,
+      "grad_norm": 2.4520591326987495,
+      "learning_rate": 1.999992660921826e-05,
+      "loss": 2.2326,
+      "step": 42
+    },
+    {
+      "epoch": 0.10225921521997622,
+      "grad_norm": 1.4817926319377914,
+      "learning_rate": 1.999983487099356e-05,
+      "loss": 2.541,
+      "step": 43
+    },
+    {
+      "epoch": 0.10463733650416171,
+      "grad_norm": 2.4022755616863956,
+      "learning_rate": 1.999970643795027e-05,
+      "loss": 2.3645,
+      "step": 44
+    },
+    {
+      "epoch": 0.1070154577883472,
+      "grad_norm": 2.476073007712477,
+      "learning_rate": 1.9999541310559686e-05,
+      "loss": 2.5051,
+      "step": 45
+    },
+    {
+      "epoch": 0.10939357907253269,
+      "grad_norm": 3.428213096316913,
+      "learning_rate": 1.9999339489427746e-05,
+      "loss": 2.3605,
+      "step": 46
+    },
+    {
+      "epoch": 0.1117717003567182,
+      "grad_norm": 2.464783346708793,
+      "learning_rate": 1.9999100975295046e-05,
+      "loss": 2.3785,
+      "step": 47
+    },
+    {
+      "epoch": 0.11414982164090369,
+      "grad_norm": 2.0686444585541754,
+      "learning_rate": 1.999882576903682e-05,
+      "loss": 2.6035,
+      "step": 48
+    },
+    {
+      "epoch": 0.11652794292508918,
+      "grad_norm": 2.30832669621963,
+      "learning_rate": 1.9998513871662945e-05,
+      "loss": 2.3982,
+      "step": 49
+    },
+    {
+      "epoch": 0.11890606420927467,
+      "grad_norm": 1.7443884357427357,
+      "learning_rate": 1.9998165284317944e-05,
+      "loss": 2.2344,
+      "step": 50
+    },
+    {
+      "epoch": 0.12128418549346016,
+      "grad_norm": 1.862498116501275,
+      "learning_rate": 1.999778000828098e-05,
+      "loss": 2.0631,
+      "step": 51
+    },
+    {
+      "epoch": 0.12366230677764566,
+      "grad_norm": 1.3054707043181313,
+      "learning_rate": 1.9997358044965833e-05,
+      "loss": 2.4434,
+      "step": 52
+    },
+    {
+      "epoch": 0.12604042806183116,
+      "grad_norm": 2.6645481874919583,
+      "learning_rate": 1.9996899395920915e-05,
+      "loss": 2.4647,
+      "step": 53
+    },
+    {
+      "epoch": 0.12841854934601665,
+      "grad_norm": 1.6884769054479885,
+      "learning_rate": 1.999640406282926e-05,
+      "loss": 2.2864,
+      "step": 54
+    },
+    {
+      "epoch": 0.13079667063020214,
+      "grad_norm": 2.281038989939936,
+      "learning_rate": 1.9995872047508516e-05,
+      "loss": 2.2999,
+      "step": 55
+    },
+    {
+      "epoch": 0.13317479191438764,
+      "grad_norm": 1.8457967068582515,
+      "learning_rate": 1.9995303351910934e-05,
+      "loss": 2.5117,
+      "step": 56
+    },
+    {
+      "epoch": 0.13555291319857313,
+      "grad_norm": 1.593901860141829,
+      "learning_rate": 1.9994697978123363e-05,
+      "loss": 2.1696,
+      "step": 57
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 2.168613470386471,
+      "learning_rate": 1.9994055928367256e-05,
+      "loss": 2.6852,
+      "step": 58
+    },
+    {
+      "epoch": 0.1403091557669441,
+      "grad_norm": 1.5398971278043392,
+      "learning_rate": 1.999337720499863e-05,
+      "loss": 2.396,
+      "step": 59
+    },
+    {
+      "epoch": 0.1426872770511296,
+      "grad_norm": 1.7618365641290346,
+      "learning_rate": 1.99926618105081e-05,
+      "loss": 2.047,
+      "step": 60
+    },
+    {
+      "epoch": 0.1450653983353151,
+      "grad_norm": 1.3935889380673343,
+      "learning_rate": 1.9991909747520835e-05,
+      "loss": 2.1071,
+      "step": 61
+    },
+    {
+      "epoch": 0.14744351961950058,
+      "grad_norm": 1.3063537354837544,
+      "learning_rate": 1.999112101879656e-05,
+      "loss": 2.3992,
+      "step": 62
+    },
+    {
+      "epoch": 0.14982164090368608,
+      "grad_norm": 1.025531260937785,
+      "learning_rate": 1.9990295627229544e-05,
+      "loss": 2.4764,
+      "step": 63
+    },
+    {
+      "epoch": 0.15219976218787157,
+      "grad_norm": 1.0662434476421614,
+      "learning_rate": 1.99894335758486e-05,
+      "loss": 2.0608,
+      "step": 64
+    },
+    {
+      "epoch": 0.1545778834720571,
+      "grad_norm": 1.335312659171346,
+      "learning_rate": 1.9988534867817065e-05,
+      "loss": 2.0345,
+      "step": 65
+    },
+    {
+      "epoch": 0.15695600475624258,
+      "grad_norm": 1.7866360477276542,
+      "learning_rate": 1.9987599506432785e-05,
+      "loss": 2.4781,
+      "step": 66
+    },
+    {
+      "epoch": 0.15933412604042807,
+      "grad_norm": 1.3661388854405736,
+      "learning_rate": 1.9986627495128105e-05,
+      "loss": 2.3895,
+      "step": 67
+    },
+    {
+      "epoch": 0.16171224732461356,
+      "grad_norm": 1.69413799763372,
+      "learning_rate": 1.9985618837469864e-05,
+      "loss": 2.415,
+      "step": 68
+    },
+    {
+      "epoch": 0.16409036860879905,
+      "grad_norm": 1.6458921632697698,
+      "learning_rate": 1.998457353715938e-05,
+      "loss": 2.4017,
+      "step": 69
+    },
+    {
+      "epoch": 0.16646848989298454,
+      "grad_norm": 1.705150979387567,
+      "learning_rate": 1.998349159803241e-05,
+      "loss": 2.317,
+      "step": 70
+    },
+    {
+      "epoch": 0.16884661117717004,
+      "grad_norm": 1.4786622132550975,
+      "learning_rate": 1.9982373024059195e-05,
+      "loss": 2.4046,
+      "step": 71
+    },
+    {
+      "epoch": 0.17122473246135553,
+      "grad_norm": 1.1257378401253821,
+      "learning_rate": 1.998121781934438e-05,
+      "loss": 2.196,
+      "step": 72
+    },
+    {
+      "epoch": 0.17360285374554102,
+      "grad_norm": 1.2450243917414825,
+      "learning_rate": 1.9980025988127037e-05,
+      "loss": 2.1235,
+      "step": 73
+    },
+    {
+      "epoch": 0.1759809750297265,
+      "grad_norm": 1.2574977788035384,
+      "learning_rate": 1.9978797534780646e-05,
+      "loss": 2.4466,
+      "step": 74
+    },
+    {
+      "epoch": 0.178359096313912,
+      "grad_norm": 1.390309850165232,
+      "learning_rate": 1.9977532463813064e-05,
+      "loss": 2.4469,
+      "step": 75
+    },
+    {
+      "epoch": 0.1807372175980975,
+      "grad_norm": 1.2146735833909619,
+      "learning_rate": 1.9976230779866527e-05,
+      "loss": 2.3705,
+      "step": 76
+    },
+    {
+      "epoch": 0.18311533888228299,
+      "grad_norm": 1.26356031715395,
+      "learning_rate": 1.9974892487717613e-05,
+      "loss": 2.4926,
+      "step": 77
+    },
+    {
+      "epoch": 0.18549346016646848,
+      "grad_norm": 1.3934309027656608,
+      "learning_rate": 1.997351759227725e-05,
+      "loss": 2.47,
+      "step": 78
+    },
+    {
+      "epoch": 0.187871581450654,
+      "grad_norm": 1.129998342751621,
+      "learning_rate": 1.9972106098590665e-05,
+      "loss": 2.3718,
+      "step": 79
+    },
+    {
+      "epoch": 0.1902497027348395,
+      "grad_norm": 1.3957720808228478,
+      "learning_rate": 1.9970658011837404e-05,
+      "loss": 2.2057,
+      "step": 80
+    },
+    {
+      "epoch": 0.19262782401902498,
+      "grad_norm": 1.5835508346410572,
+      "learning_rate": 1.9969173337331283e-05,
+      "loss": 2.2551,
+      "step": 81
+    },
+    {
+      "epoch": 0.19500594530321047,
+      "grad_norm": 1.0726363834452401,
+      "learning_rate": 1.996765208052037e-05,
+      "loss": 1.9962,
+      "step": 82
+    },
+    {
+      "epoch": 0.19738406658739596,
+      "grad_norm": 1.5380248440103288,
+      "learning_rate": 1.9966094246986983e-05,
+      "loss": 2.1986,
+      "step": 83
+    },
+    {
+      "epoch": 0.19976218787158145,
+      "grad_norm": 1.2953387209833067,
+      "learning_rate": 1.9964499842447665e-05,
+      "loss": 2.5842,
+      "step": 84
+    },
+    {
+      "epoch": 0.20214030915576695,
+      "grad_norm": 1.0712033116668103,
+      "learning_rate": 1.9962868872753144e-05,
+      "loss": 2.1298,
+      "step": 85
+    },
+    {
+      "epoch": 0.20451843043995244,
+      "grad_norm": 1.095761902776689,
+      "learning_rate": 1.996120134388834e-05,
+      "loss": 2.3641,
+      "step": 86
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 1.186679631328553,
+      "learning_rate": 1.995949726197231e-05,
+      "loss": 2.4801,
+      "step": 87
+    },
+    {
+      "epoch": 0.20927467300832342,
+      "grad_norm": 1.1887498108170933,
+      "learning_rate": 1.9957756633258264e-05,
+      "loss": 2.2866,
+      "step": 88
+    },
+    {
+      "epoch": 0.2116527942925089,
+      "grad_norm": 1.0909023774872124,
+      "learning_rate": 1.9955979464133515e-05,
+      "loss": 2.2916,
+      "step": 89
+    },
+    {
+      "epoch": 0.2140309155766944,
+      "grad_norm": 1.1077175360558418,
+      "learning_rate": 1.995416576111945e-05,
+      "loss": 2.1077,
+      "step": 90
+    },
+    {
+      "epoch": 0.2164090368608799,
+      "grad_norm": 1.1526064687436712,
+      "learning_rate": 1.9952315530871537e-05,
+      "loss": 2.2723,
+      "step": 91
+    },
+    {
+      "epoch": 0.21878715814506539,
+      "grad_norm": 1.1818210038912647,
+      "learning_rate": 1.9950428780179274e-05,
+      "loss": 2.2338,
+      "step": 92
+    },
+    {
+      "epoch": 0.2211652794292509,
+      "grad_norm": 1.164942154271255,
+      "learning_rate": 1.994850551596617e-05,
+      "loss": 2.3817,
+      "step": 93
+    },
+    {
+      "epoch": 0.2235434007134364,
+      "grad_norm": 1.339398993177121,
+      "learning_rate": 1.9946545745289727e-05,
+      "loss": 2.5508,
+      "step": 94
+    },
+    {
+      "epoch": 0.2259215219976219,
+      "grad_norm": 1.3267763052855093,
+      "learning_rate": 1.9944549475341404e-05,
+      "loss": 2.247,
+      "step": 95
+    },
+    {
+      "epoch": 0.22829964328180738,
+      "grad_norm": 1.1953250811556597,
+      "learning_rate": 1.99425167134466e-05,
+      "loss": 2.3373,
+      "step": 96
+    },
+    {
+      "epoch": 0.23067776456599287,
+      "grad_norm": 1.4321452409301854,
+      "learning_rate": 1.9940447467064624e-05,
+      "loss": 2.4776,
+      "step": 97
+    },
+    {
+      "epoch": 0.23305588585017836,
+      "grad_norm": 1.0224444212683161,
+      "learning_rate": 1.9938341743788658e-05,
+      "loss": 2.1837,
+      "step": 98
+    },
+    {
+      "epoch": 0.23543400713436385,
+      "grad_norm": 1.0977574950238398,
+      "learning_rate": 1.9936199551345744e-05,
+      "loss": 2.1478,
+      "step": 99
+    },
+    {
+      "epoch": 0.23781212841854935,
+      "grad_norm": 1.0660069054078747,
+      "learning_rate": 1.9934020897596752e-05,
+      "loss": 2.2816,
+      "step": 100
+    },
+    {
+      "epoch": 0.24019024970273484,
+      "grad_norm": 1.0312993672336248,
+      "learning_rate": 1.9931805790536342e-05,
+      "loss": 2.2468,
+      "step": 101
+    },
+    {
+      "epoch": 0.24256837098692033,
+      "grad_norm": 1.1278898252066067,
+      "learning_rate": 1.9929554238292944e-05,
+      "loss": 2.0526,
+      "step": 102
+    },
+    {
+      "epoch": 0.24494649227110582,
+      "grad_norm": 1.288343002944789,
+      "learning_rate": 1.992726624912872e-05,
+      "loss": 2.512,
+      "step": 103
+    },
+    {
+      "epoch": 0.2473246135552913,
+      "grad_norm": 1.1840020386119305,
+      "learning_rate": 1.992494183143955e-05,
+      "loss": 2.6356,
+      "step": 104
+    },
+    {
+      "epoch": 0.2497027348394768,
+      "grad_norm": 1.1014678408276726,
+      "learning_rate": 1.9922580993754985e-05,
+      "loss": 2.3521,
+      "step": 105
+    },
+    {
+      "epoch": 0.2497027348394768,
+      "eval_loss": 2.5365779399871826,
+      "eval_runtime": 66.0796,
+      "eval_samples_per_second": 9.67,
+      "eval_steps_per_second": 1.211,
+      "step": 105
+    },
+    {
+      "epoch": 0.2520808561236623,
+      "grad_norm": 0.9957844757920508,
+      "learning_rate": 1.9920183744738208e-05,
+      "loss": 2.355,
+      "step": 106
+    },
+    {
+      "epoch": 0.2544589774078478,
+      "grad_norm": 1.0070598447313825,
+      "learning_rate": 1.9917750093186036e-05,
+      "loss": 2.1747,
+      "step": 107
+    },
+    {
+      "epoch": 0.2568370986920333,
+      "grad_norm": 1.2463453868295562,
+      "learning_rate": 1.9915280048028853e-05,
+      "loss": 2.4131,
+      "step": 108
+    },
+    {
+      "epoch": 0.25921521997621877,
+      "grad_norm": 1.158558292534161,
+      "learning_rate": 1.9912773618330595e-05,
+      "loss": 2.4527,
+      "step": 109
+    },
+    {
+      "epoch": 0.2615933412604043,
+      "grad_norm": 1.1875643459332377,
+      "learning_rate": 1.9910230813288713e-05,
+      "loss": 2.1523,
+      "step": 110
+    },
+    {
+      "epoch": 0.26397146254458975,
+      "grad_norm": 0.892269173897758,
+      "learning_rate": 1.9907651642234138e-05,
+      "loss": 1.9606,
+      "step": 111
+    },
+    {
+      "epoch": 0.26634958382877527,
+      "grad_norm": 1.181952902180908,
+      "learning_rate": 1.9905036114631247e-05,
+      "loss": 2.3201,
+      "step": 112
+    },
+    {
+      "epoch": 0.26872770511296074,
+      "grad_norm": 0.9689153704257877,
+      "learning_rate": 1.990238424007783e-05,
+      "loss": 2.2329,
+      "step": 113
+    },
+    {
+      "epoch": 0.27110582639714625,
+      "grad_norm": 1.3665918769424286,
+      "learning_rate": 1.989969602830505e-05,
+      "loss": 2.2387,
+      "step": 114
+    },
+    {
+      "epoch": 0.2734839476813318,
+      "grad_norm": 1.0478434719151144,
+      "learning_rate": 1.9896971489177417e-05,
+      "loss": 2.2798,
+      "step": 115
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 1.5752154316391798,
+      "learning_rate": 1.9894210632692745e-05,
+      "loss": 2.201,
+      "step": 116
+    },
+    {
+      "epoch": 0.27824019024970276,
+      "grad_norm": 1.0264277011384757,
+      "learning_rate": 1.9891413468982112e-05,
+      "loss": 2.2756,
+      "step": 117
+    },
+    {
+      "epoch": 0.2806183115338882,
+      "grad_norm": 1.1750703393359614,
+      "learning_rate": 1.988858000830983e-05,
+      "loss": 2.1907,
+      "step": 118
+    },
+    {
+      "epoch": 0.28299643281807374,
+      "grad_norm": 0.9456957190962577,
+      "learning_rate": 1.9885710261073402e-05,
+      "loss": 2.2993,
+      "step": 119
+    },
+    {
+      "epoch": 0.2853745541022592,
+      "grad_norm": 1.37591692336223,
+      "learning_rate": 1.9882804237803487e-05,
+      "loss": 2.0751,
+      "step": 120
+    },
+    {
+      "epoch": 0.2877526753864447,
+      "grad_norm": 0.991181263305241,
+      "learning_rate": 1.9879861949163863e-05,
+      "loss": 2.1946,
+      "step": 121
+    },
+    {
+      "epoch": 0.2901307966706302,
+      "grad_norm": 1.2826616603092615,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 2.3084,
+      "step": 122
+    },
+    {
+      "epoch": 0.2925089179548157,
+      "grad_norm": 1.3162982027829009,
+      "learning_rate": 1.987386861909593e-05,
+      "loss": 2.294,
+      "step": 123
+    },
+    {
+      "epoch": 0.29488703923900117,
+      "grad_norm": 1.086311999313279,
+      "learning_rate": 1.98708175996604e-05,
+      "loss": 2.3025,
+      "step": 124
+    },
+    {
+      "epoch": 0.2972651605231867,
+      "grad_norm": 1.10683170372015,
+      "learning_rate": 1.986773035884064e-05,
+      "loss": 2.3447,
+      "step": 125
+    },
+    {
+      "epoch": 0.29964328180737215,
+      "grad_norm": 1.090568761480393,
+      "learning_rate": 1.9864606907965407e-05,
+      "loss": 2.4104,
+      "step": 126
+    },
+    {
+      "epoch": 0.30202140309155767,
+      "grad_norm": 1.4024759238343605,
+      "learning_rate": 1.986144725849634e-05,
+      "loss": 2.298,
+      "step": 127
+    },
+    {
+      "epoch": 0.30439952437574314,
+      "grad_norm": 0.9324914520062791,
+      "learning_rate": 1.9858251422027903e-05,
+      "loss": 2.1123,
+      "step": 128
+    },
+    {
+      "epoch": 0.30677764565992865,
+      "grad_norm": 1.3818136151492852,
+      "learning_rate": 1.9855019410287355e-05,
+      "loss": 2.2786,
+      "step": 129
+    },
+    {
+      "epoch": 0.3091557669441142,
+      "grad_norm": 0.9879756737720099,
+      "learning_rate": 1.98517512351347e-05,
+      "loss": 2.2735,
+      "step": 130
+    },
+    {
+      "epoch": 0.31153388822829964,
+      "grad_norm": 1.4107106057474024,
+      "learning_rate": 1.9848446908562647e-05,
+      "loss": 2.2421,
+      "step": 131
+    },
+    {
+      "epoch": 0.31391200951248516,
+      "grad_norm": 0.978862094447652,
+      "learning_rate": 1.9845106442696563e-05,
+      "loss": 2.4152,
+      "step": 132
+    },
+    {
+      "epoch": 0.3162901307966706,
+      "grad_norm": 1.3714074038447606,
+      "learning_rate": 1.9841729849794427e-05,
+      "loss": 2.4567,
+      "step": 133
+    },
+    {
+      "epoch": 0.31866825208085614,
+      "grad_norm": 1.030641093673837,
+      "learning_rate": 1.983831714224679e-05,
+      "loss": 2.3015,
+      "step": 134
+    },
+    {
+      "epoch": 0.3210463733650416,
+      "grad_norm": 1.1744699755999302,
+      "learning_rate": 1.9834868332576727e-05,
+      "loss": 2.2878,
+      "step": 135
+    },
+    {
+      "epoch": 0.3234244946492271,
+      "grad_norm": 0.9733999816490441,
+      "learning_rate": 1.9831383433439798e-05,
+      "loss": 2.1571,
+      "step": 136
+    },
+    {
+      "epoch": 0.3258026159334126,
+      "grad_norm": 1.0470367999253474,
+      "learning_rate": 1.982786245762398e-05,
+      "loss": 2.0943,
+      "step": 137
+    },
+    {
+      "epoch": 0.3281807372175981,
+      "grad_norm": 1.0748276455064096,
+      "learning_rate": 1.9824305418049645e-05,
+      "loss": 2.4156,
+      "step": 138
+    },
+    {
+      "epoch": 0.33055885850178357,
+      "grad_norm": 1.0220509349947084,
+      "learning_rate": 1.9820712327769503e-05,
+      "loss": 2.1898,
+      "step": 139
+    },
+    {
+      "epoch": 0.3329369797859691,
+      "grad_norm": 0.9811166423920332,
+      "learning_rate": 1.9817083199968552e-05,
+      "loss": 2.3449,
+      "step": 140
+    },
+    {
+      "epoch": 0.33531510107015455,
+      "grad_norm": 1.0664757695722766,
+      "learning_rate": 1.9813418047964025e-05,
+      "loss": 2.1514,
+      "step": 141
+    },
+    {
+      "epoch": 0.3376932223543401,
+      "grad_norm": 1.1228830278366924,
+      "learning_rate": 1.9809716885205363e-05,
+      "loss": 2.3371,
+      "step": 142
+    },
+    {
+      "epoch": 0.3400713436385256,
+      "grad_norm": 1.0703957613617774,
+      "learning_rate": 1.980597972527413e-05,
+      "loss": 2.2577,
+      "step": 143
+    },
+    {
+      "epoch": 0.34244946492271106,
+      "grad_norm": 0.9971842999532138,
+      "learning_rate": 1.9802206581883992e-05,
+      "loss": 2.2048,
+      "step": 144
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 0.9969712850303254,
+      "learning_rate": 1.979839746888067e-05,
+      "loss": 2.1725,
+      "step": 145
+    },
+    {
+      "epoch": 0.34720570749108204,
+      "grad_norm": 0.9782490093980141,
+      "learning_rate": 1.979455240024186e-05,
+      "loss": 2.1598,
+      "step": 146
+    },
+    {
+      "epoch": 0.34958382877526756,
+      "grad_norm": 1.1595035293528873,
+      "learning_rate": 1.97906713900772e-05,
+      "loss": 2.1812,
+      "step": 147
+    },
+    {
+      "epoch": 0.351961950059453,
+      "grad_norm": 1.0488323565717943,
+      "learning_rate": 1.9786754452628226e-05,
+      "loss": 2.126,
+      "step": 148
+    },
+    {
+      "epoch": 0.35434007134363854,
+      "grad_norm": 1.0236205683546673,
+      "learning_rate": 1.9782801602268306e-05,
+      "loss": 1.9399,
+      "step": 149
+    },
+    {
+      "epoch": 0.356718192627824,
+      "grad_norm": 0.983049547537296,
+      "learning_rate": 1.9778812853502592e-05,
+      "loss": 2.0336,
+      "step": 150
+    },
+    {
+      "epoch": 0.3590963139120095,
+      "grad_norm": 1.0856474713800959,
+      "learning_rate": 1.9774788220967968e-05,
+      "loss": 2.2103,
+      "step": 151
+    },
+    {
+      "epoch": 0.361474435196195,
+      "grad_norm": 1.098143269144179,
+      "learning_rate": 1.9770727719432994e-05,
+      "loss": 2.1425,
+      "step": 152
+    },
+    {
+      "epoch": 0.3638525564803805,
+      "grad_norm": 1.1908904777112574,
+      "learning_rate": 1.9766631363797852e-05,
+      "loss": 2.2516,
+      "step": 153
+    },
+    {
+      "epoch": 0.36623067776456597,
+      "grad_norm": 1.1823343263781934,
+      "learning_rate": 1.9762499169094288e-05,
+      "loss": 2.0991,
+      "step": 154
+    },
+    {
+      "epoch": 0.3686087990487515,
+      "grad_norm": 1.1543274307271654,
+      "learning_rate": 1.9758331150485576e-05,
+      "loss": 2.2917,
+      "step": 155
+    },
+    {
+      "epoch": 0.37098692033293695,
+      "grad_norm": 1.1828452156246019,
+      "learning_rate": 1.9754127323266426e-05,
+      "loss": 2.3577,
+      "step": 156
+    },
+    {
+      "epoch": 0.3733650416171225,
+      "grad_norm": 1.2458434785978698,
+      "learning_rate": 1.9749887702862972e-05,
+      "loss": 2.2291,
+      "step": 157
+    },
+    {
+      "epoch": 0.375743162901308,
+      "grad_norm": 1.0632348458757013,
+      "learning_rate": 1.9745612304832672e-05,
+      "loss": 2.495,
+      "step": 158
+    },
+    {
+      "epoch": 0.37812128418549346,
+      "grad_norm": 1.2413557275846534,
+      "learning_rate": 1.9741301144864284e-05,
+      "loss": 2.3006,
+      "step": 159
+    },
+    {
+      "epoch": 0.380499405469679,
+      "grad_norm": 1.068837985332943,
+      "learning_rate": 1.9736954238777793e-05,
+      "loss": 2.2228,
+      "step": 160
+    },
+    {
+      "epoch": 0.38287752675386444,
+      "grad_norm": 1.181973772137545,
+      "learning_rate": 1.9732571602524353e-05,
+      "loss": 2.3419,
+      "step": 161
+    },
+    {
+      "epoch": 0.38525564803804996,
+      "grad_norm": 0.9361759344356807,
+      "learning_rate": 1.972815325218624e-05,
+      "loss": 2.2727,
+      "step": 162
+    },
+    {
+      "epoch": 0.3876337693222354,
+      "grad_norm": 1.2300672941710984,
+      "learning_rate": 1.9723699203976768e-05,
+      "loss": 2.3947,
+      "step": 163
+    },
+    {
+      "epoch": 0.39001189060642094,
+      "grad_norm": 0.9647921025871186,
+      "learning_rate": 1.9719209474240263e-05,
+      "loss": 1.8388,
+      "step": 164
+    },
+    {
+      "epoch": 0.3923900118906064,
+      "grad_norm": 1.1390311715526416,
+      "learning_rate": 1.971468407945198e-05,
+      "loss": 2.4054,
+      "step": 165
+    },
+    {
+      "epoch": 0.3947681331747919,
+      "grad_norm": 0.9830051867519547,
+      "learning_rate": 1.9710123036218044e-05,
+      "loss": 2.0355,
+      "step": 166
+    },
+    {
+      "epoch": 0.3971462544589774,
+      "grad_norm": 1.1244517585073737,
+      "learning_rate": 1.97055263612754e-05,
+      "loss": 2.0188,
+      "step": 167
+    },
+    {
+      "epoch": 0.3995243757431629,
+      "grad_norm": 1.0256020852263494,
+      "learning_rate": 1.9700894071491736e-05,
+      "loss": 2.0774,
+      "step": 168
+    },
+    {
+      "epoch": 0.40190249702734837,
+      "grad_norm": 1.011023720252716,
+      "learning_rate": 1.9696226183865436e-05,
+      "loss": 2.2592,
+      "step": 169
+    },
+    {
+      "epoch": 0.4042806183115339,
+      "grad_norm": 1.046975898884085,
+      "learning_rate": 1.969152271552552e-05,
+      "loss": 2.1791,
+      "step": 170
+    },
+    {
+      "epoch": 0.40665873959571935,
+      "grad_norm": 1.1800984480399852,
+      "learning_rate": 1.9686783683731557e-05,
+      "loss": 2.3941,
+      "step": 171
+    },
+    {
+      "epoch": 0.4090368608799049,
+      "grad_norm": 1.2459882622321672,
+      "learning_rate": 1.9682009105873633e-05,
+      "loss": 2.1522,
+      "step": 172
+    },
+    {
+      "epoch": 0.4114149821640904,
+      "grad_norm": 1.0732133381850257,
+      "learning_rate": 1.9677198999472257e-05,
+      "loss": 2.1233,
+      "step": 173
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 1.2405484917580802,
+      "learning_rate": 1.967235338217832e-05,
+      "loss": 2.3016,
+      "step": 174
+    },
+    {
+      "epoch": 0.4161712247324614,
+      "grad_norm": 1.0759940201219593,
+      "learning_rate": 1.9667472271773026e-05,
+      "loss": 2.2947,
+      "step": 175
+    },
+    {
+      "epoch": 0.41854934601664684,
+      "grad_norm": 1.2008734320661734,
+      "learning_rate": 1.9662555686167808e-05,
+      "loss": 2.2155,
+      "step": 176
+    },
+    {
+      "epoch": 0.42092746730083236,
+      "grad_norm": 0.9303619935178572,
+      "learning_rate": 1.965760364340429e-05,
+      "loss": 2.1234,
+      "step": 177
+    },
+    {
+      "epoch": 0.4233055885850178,
+      "grad_norm": 1.3884826767438652,
+      "learning_rate": 1.9652616161654204e-05,
+      "loss": 2.2539,
+      "step": 178
+    },
+    {
+      "epoch": 0.42568370986920334,
+      "grad_norm": 0.9947187673832885,
+      "learning_rate": 1.9647593259219328e-05,
+      "loss": 2.2052,
+      "step": 179
+    },
+    {
+      "epoch": 0.4280618311533888,
+      "grad_norm": 1.4655922792083054,
+      "learning_rate": 1.964253495453141e-05,
+      "loss": 2.1552,
+      "step": 180
+    },
+    {
+      "epoch": 0.4304399524375743,
+      "grad_norm": 1.1481294188693778,
+      "learning_rate": 1.963744126615212e-05,
+      "loss": 2.3942,
+      "step": 181
+    },
+    {
+      "epoch": 0.4328180737217598,
+      "grad_norm": 1.239760521409481,
+      "learning_rate": 1.9632312212772956e-05,
+      "loss": 2.3091,
+      "step": 182
+    },
+    {
+      "epoch": 0.4351961950059453,
+      "grad_norm": 1.0524654460411744,
+      "learning_rate": 1.9627147813215207e-05,
+      "loss": 2.302,
+      "step": 183
+    },
+    {
+      "epoch": 0.43757431629013077,
+      "grad_norm": 1.0231645108607732,
+      "learning_rate": 1.9621948086429847e-05,
+      "loss": 2.2334,
+      "step": 184
+    },
+    {
+      "epoch": 0.4399524375743163,
+      "grad_norm": 1.0600582051447691,
+      "learning_rate": 1.9616713051497496e-05,
+      "loss": 2.2044,
+      "step": 185
+    },
+    {
+      "epoch": 0.4423305588585018,
+      "grad_norm": 1.0861978175484295,
+      "learning_rate": 1.9611442727628344e-05,
+      "loss": 2.3267,
+      "step": 186
+    },
+    {
+      "epoch": 0.4447086801426873,
+      "grad_norm": 1.0122924353396487,
+      "learning_rate": 1.960613713416206e-05,
+      "loss": 2.2327,
+      "step": 187
+    },
+    {
+      "epoch": 0.4470868014268728,
+      "grad_norm": 1.1275635495135592,
+      "learning_rate": 1.9600796290567747e-05,
+      "loss": 2.2474,
+      "step": 188
+    },
+    {
+      "epoch": 0.44946492271105826,
+      "grad_norm": 1.0778906611663819,
+      "learning_rate": 1.9595420216443864e-05,
+      "loss": 2.2777,
+      "step": 189
+    },
+    {
+      "epoch": 0.4518430439952438,
+      "grad_norm": 1.0593499669893551,
+      "learning_rate": 1.9590008931518133e-05,
+      "loss": 2.4937,
+      "step": 190
+    },
+    {
+      "epoch": 0.45422116527942924,
+      "grad_norm": 1.0887914371115388,
+      "learning_rate": 1.9584562455647494e-05,
+      "loss": 2.2577,
+      "step": 191
+    },
+    {
+      "epoch": 0.45659928656361476,
+      "grad_norm": 1.0280779311785984,
+      "learning_rate": 1.9579080808818035e-05,
+      "loss": 2.2352,
+      "step": 192
+    },
+    {
+      "epoch": 0.4589774078478002,
+      "grad_norm": 1.1201705856067985,
+      "learning_rate": 1.9573564011144873e-05,
+      "loss": 2.1482,
+      "step": 193
+    },
+    {
+      "epoch": 0.46135552913198574,
+      "grad_norm": 1.0039435227655624,
+      "learning_rate": 1.9568012082872148e-05,
+      "loss": 2.1069,
+      "step": 194
+    },
+    {
+      "epoch": 0.4637336504161712,
+      "grad_norm": 1.0523831000821406,
+      "learning_rate": 1.9562425044372884e-05,
+      "loss": 1.9268,
+      "step": 195
+    },
+    {
+      "epoch": 0.4661117717003567,
+      "grad_norm": 1.0635880350342213,
+      "learning_rate": 1.9556802916148963e-05,
+      "loss": 2.2722,
+      "step": 196
+    },
+    {
+      "epoch": 0.4684898929845422,
+      "grad_norm": 2.4351848601787287,
+      "learning_rate": 1.955114571883102e-05,
+      "loss": 2.1402,
+      "step": 197
+    },
+    {
+      "epoch": 0.4708680142687277,
+      "grad_norm": 1.2199308274597462,
+      "learning_rate": 1.9545453473178384e-05,
+      "loss": 2.2599,
+      "step": 198
+    },
+    {
+      "epoch": 0.47324613555291317,
+      "grad_norm": 0.9936114796299212,
+      "learning_rate": 1.9539726200078987e-05,
+      "loss": 2.0662,
+      "step": 199
+    },
+    {
+      "epoch": 0.4756242568370987,
+      "grad_norm": 1.0692703333507547,
+      "learning_rate": 1.9533963920549307e-05,
+      "loss": 2.3739,
+      "step": 200
+    },
+    {
+      "epoch": 0.4780023781212842,
+      "grad_norm": 1.0406002686664542,
+      "learning_rate": 1.9528166655734267e-05,
+      "loss": 2.3611,
+      "step": 201
+    },
+    {
+      "epoch": 0.4803804994054697,
+      "grad_norm": 1.9375905536343168,
+      "learning_rate": 1.9522334426907185e-05,
+      "loss": 2.0971,
+      "step": 202
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 1.024548704059581,
+      "learning_rate": 1.951646725546966e-05,
+      "loss": 2.2498,
+      "step": 203
+    },
+    {
+      "epoch": 0.48513674197384066,
+      "grad_norm": 1.0033895284405978,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 2.299,
+      "step": 204
+    },
+    {
+      "epoch": 0.4875148632580262,
+      "grad_norm": 1.541631519071697,
+      "learning_rate": 1.950462817101079e-05,
+      "loss": 2.4076,
+      "step": 205
+    },
+    {
+      "epoch": 0.48989298454221164,
+      "grad_norm": 0.9499702987331401,
+      "learning_rate": 1.9498656301433466e-05,
+      "loss": 2.0754,
+      "step": 206
+    },
+    {
+      "epoch": 0.49227110582639716,
+      "grad_norm": 1.099383371761328,
+      "learning_rate": 1.9492649576133594e-05,
+      "loss": 2.2514,
+      "step": 207
+    },
+    {
+      "epoch": 0.4946492271105826,
+      "grad_norm": 0.9296431838496088,
+      "learning_rate": 1.94866080171531e-05,
+      "loss": 2.2308,
+      "step": 208
+    },
+    {
+      "epoch": 0.49702734839476814,
+      "grad_norm": 4.140796209905845,
+      "learning_rate": 1.9480531646661753e-05,
+      "loss": 2.4388,
+      "step": 209
+    },
+    {
+      "epoch": 0.4994054696789536,
+      "grad_norm": 1.011142238194789,
+      "learning_rate": 1.9474420486957045e-05,
+      "loss": 2.2414,
+      "step": 210
+    },
+    {
+      "epoch": 0.4994054696789536,
+      "eval_loss": 2.49302339553833,
+      "eval_runtime": 65.8636,
+      "eval_samples_per_second": 9.702,
+      "eval_steps_per_second": 1.215,
+      "step": 210
+    },
+    {
+      "epoch": 0.5017835909631391,
+      "grad_norm": 0.9941401750604694,
+      "learning_rate": 1.9468274560464134e-05,
+      "loss": 2.2182,
+      "step": 211
+    },
+    {
+      "epoch": 0.5041617122473246,
+      "grad_norm": 1.035068690961865,
+      "learning_rate": 1.9462093889735766e-05,
+      "loss": 2.3569,
+      "step": 212
+    },
+    {
+      "epoch": 0.5065398335315101,
+      "grad_norm": 1.5052185888965133,
+      "learning_rate": 1.945587849745217e-05,
+      "loss": 2.4474,
+      "step": 213
+    },
+    {
+      "epoch": 0.5089179548156956,
+      "grad_norm": 1.01730612362564,
+      "learning_rate": 1.944962840642099e-05,
+      "loss": 2.3912,
+      "step": 214
+    },
+    {
+      "epoch": 0.5112960760998811,
+      "grad_norm": 0.9513743221196854,
+      "learning_rate": 1.9443343639577206e-05,
+      "loss": 2.2842,
+      "step": 215
+    },
+    {
+      "epoch": 0.5136741973840666,
+      "grad_norm": 0.9759286606572132,
+      "learning_rate": 1.943702421998303e-05,
+      "loss": 2.3117,
+      "step": 216
+    },
+    {
+      "epoch": 0.5160523186682521,
+      "grad_norm": 0.9357816333722543,
+      "learning_rate": 1.9430670170827844e-05,
+      "loss": 2.1091,
+      "step": 217
+    },
+    {
+      "epoch": 0.5184304399524375,
+      "grad_norm": 1.1804547022456764,
+      "learning_rate": 1.94242815154281e-05,
+      "loss": 2.2333,
+      "step": 218
+    },
+    {
+      "epoch": 0.5208085612366231,
+      "grad_norm": 1.0289463651234612,
+      "learning_rate": 1.9417858277227244e-05,
+      "loss": 2.0495,
+      "step": 219
+    },
+    {
+      "epoch": 0.5231866825208086,
+      "grad_norm": 0.9297621773216459,
+      "learning_rate": 1.9411400479795618e-05,
+      "loss": 2.2102,
+      "step": 220
+    },
+    {
+      "epoch": 0.525564803804994,
+      "grad_norm": 0.9468558483075745,
+      "learning_rate": 1.9404908146830383e-05,
+      "loss": 2.2593,
+      "step": 221
+    },
+    {
+      "epoch": 0.5279429250891795,
+      "grad_norm": 1.03418169808611,
+      "learning_rate": 1.9398381302155435e-05,
+      "loss": 2.185,
+      "step": 222
+    },
+    {
+      "epoch": 0.5303210463733651,
+      "grad_norm": 0.84914434823741,
+      "learning_rate": 1.93918199697213e-05,
+      "loss": 1.8499,
+      "step": 223
+    },
+    {
+      "epoch": 0.5326991676575505,
+      "grad_norm": 1.0824133835143805,
+      "learning_rate": 1.9385224173605072e-05,
+      "loss": 2.2171,
+      "step": 224
+    },
+    {
+      "epoch": 0.535077288941736,
+      "grad_norm": 0.9478237708740055,
+      "learning_rate": 1.9378593938010302e-05,
+      "loss": 2.2397,
+      "step": 225
+    },
+    {
+      "epoch": 0.5374554102259215,
+      "grad_norm": 2.4277055664527065,
+      "learning_rate": 1.937192928726692e-05,
+      "loss": 2.2262,
+      "step": 226
+    },
+    {
+      "epoch": 0.539833531510107,
+      "grad_norm": 1.5223794932859396,
+      "learning_rate": 1.936523024583115e-05,
+      "loss": 2.2664,
+      "step": 227
+    },
+    {
+      "epoch": 0.5422116527942925,
+      "grad_norm": 1.2991334808397113,
+      "learning_rate": 1.9358496838285408e-05,
+      "loss": 2.1967,
+      "step": 228
+    },
+    {
+      "epoch": 0.544589774078478,
+      "grad_norm": 1.132823106642245,
+      "learning_rate": 1.9351729089338214e-05,
+      "loss": 2.2516,
+      "step": 229
+    },
+    {
+      "epoch": 0.5469678953626635,
+      "grad_norm": 1.0872118899678849,
+      "learning_rate": 1.9344927023824112e-05,
+      "loss": 2.3602,
+      "step": 230
+    },
+    {
+      "epoch": 0.549346016646849,
+      "grad_norm": 0.9596136441748024,
+      "learning_rate": 1.933809066670357e-05,
+      "loss": 2.3944,
+      "step": 231
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 1.3340981902899025,
+      "learning_rate": 1.9331220043062894e-05,
+      "loss": 1.8085,
+      "step": 232
+    },
+    {
+      "epoch": 0.5541022592152199,
+      "grad_norm": 0.9780209403031624,
+      "learning_rate": 1.9324315178114127e-05,
+      "loss": 2.2626,
+      "step": 233
+    },
+    {
+      "epoch": 0.5564803804994055,
+      "grad_norm": 0.8441113781121591,
+      "learning_rate": 1.9317376097194964e-05,
+      "loss": 1.904,
+      "step": 234
+    },
+    {
+      "epoch": 0.558858501783591,
+      "grad_norm": 0.9000708142025062,
+      "learning_rate": 1.9310402825768655e-05,
+      "loss": 2.1239,
+      "step": 235
+    },
+    {
+      "epoch": 0.5612366230677764,
+      "grad_norm": 0.9819927724537073,
+      "learning_rate": 1.9303395389423918e-05,
+      "loss": 2.2526,
+      "step": 236
+    },
+    {
+      "epoch": 0.5636147443519619,
+      "grad_norm": 0.9835396213176555,
+      "learning_rate": 1.9296353813874838e-05,
+      "loss": 2.1293,
+      "step": 237
+    },
+    {
+      "epoch": 0.5659928656361475,
+      "grad_norm": 0.8705520689865173,
+      "learning_rate": 1.9289278124960777e-05,
+      "loss": 1.9911,
+      "step": 238
+    },
+    {
+      "epoch": 0.5683709869203329,
+      "grad_norm": 1.0080028925541162,
+      "learning_rate": 1.9282168348646268e-05,
+      "loss": 2.0903,
+      "step": 239
+    },
+    {
+      "epoch": 0.5707491082045184,
+      "grad_norm": 1.0304202570689127,
+      "learning_rate": 1.927502451102095e-05,
+      "loss": 2.0784,
+      "step": 240
+    },
+    {
+      "epoch": 0.5731272294887039,
+      "grad_norm": 1.0836386452615874,
+      "learning_rate": 1.926784663829943e-05,
+      "loss": 2.2297,
+      "step": 241
+    },
+    {
+      "epoch": 0.5755053507728894,
+      "grad_norm": 0.972327672270083,
+      "learning_rate": 1.926063475682121e-05,
+      "loss": 2.2518,
+      "step": 242
+    },
+    {
+      "epoch": 0.5778834720570749,
+      "grad_norm": 1.1250553185490555,
+      "learning_rate": 1.9253388893050612e-05,
+      "loss": 2.2419,
+      "step": 243
+    },
+    {
+      "epoch": 0.5802615933412604,
+      "grad_norm": 1.6628057931903235,
+      "learning_rate": 1.924610907357663e-05,
+      "loss": 2.2409,
+      "step": 244
+    },
+    {
+      "epoch": 0.582639714625446,
+      "grad_norm": 1.3164993193299928,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 2.4277,
+      "step": 245
+    },
+    {
+      "epoch": 0.5850178359096314,
+      "grad_norm": 0.9607216662774912,
+      "learning_rate": 1.9231447674497444e-05,
+      "loss": 2.0357,
+      "step": 246
+    },
+    {
+      "epoch": 0.5873959571938169,
+      "grad_norm": 0.9753564233974246,
+      "learning_rate": 1.922406614869287e-05,
+      "loss": 2.1743,
+      "step": 247
+    },
+    {
+      "epoch": 0.5897740784780023,
+      "grad_norm": 0.9514731158004311,
+      "learning_rate": 1.9216650774785975e-05,
+      "loss": 2.2775,
+      "step": 248
+    },
+    {
+      "epoch": 0.5921521997621879,
+      "grad_norm": 1.1514073630104866,
+      "learning_rate": 1.9209201579987777e-05,
+      "loss": 2.0307,
+      "step": 249
+    },
+    {
+      "epoch": 0.5945303210463734,
+      "grad_norm": 1.0070268666833808,
+      "learning_rate": 1.9201718591633417e-05,
+      "loss": 2.405,
+      "step": 250
+    },
+    {
+      "epoch": 0.5969084423305588,
+      "grad_norm": 0.9724192667298419,
+      "learning_rate": 1.9194201837182045e-05,
+      "loss": 2.3375,
+      "step": 251
+    },
+    {
+      "epoch": 0.5992865636147443,
+      "grad_norm": 0.9184225452723915,
+      "learning_rate": 1.9186651344216703e-05,
+      "loss": 2.0576,
+      "step": 252
+    },
+    {
+      "epoch": 0.6016646848989299,
+      "grad_norm": 0.9277770597042948,
+      "learning_rate": 1.9179067140444246e-05,
+      "loss": 2.065,
+      "step": 253
+    },
+    {
+      "epoch": 0.6040428061831153,
+      "grad_norm": 0.9971241617303254,
+      "learning_rate": 1.9171449253695233e-05,
+      "loss": 2.0388,
+      "step": 254
+    },
+    {
+      "epoch": 0.6064209274673008,
+      "grad_norm": 0.968526625269074,
+      "learning_rate": 1.9163797711923822e-05,
+      "loss": 2.2154,
+      "step": 255
+    },
+    {
+      "epoch": 0.6087990487514863,
+      "grad_norm": 1.0974259939522593,
+      "learning_rate": 1.9156112543207674e-05,
+      "loss": 2.0897,
+      "step": 256
+    },
+    {
+      "epoch": 0.6111771700356718,
+      "grad_norm": 1.0661979784688387,
+      "learning_rate": 1.9148393775747842e-05,
+      "loss": 2.3986,
+      "step": 257
+    },
+    {
+      "epoch": 0.6135552913198573,
+      "grad_norm": 1.3741439020152701,
+      "learning_rate": 1.9140641437868664e-05,
+      "loss": 2.1249,
+      "step": 258
+    },
+    {
+      "epoch": 0.6159334126040428,
+      "grad_norm": 0.9207355567454468,
+      "learning_rate": 1.913285555801768e-05,
+      "loss": 2.2715,
+      "step": 259
+    },
+    {
+      "epoch": 0.6183115338882283,
+      "grad_norm": 1.0171260467381849,
+      "learning_rate": 1.9125036164765502e-05,
+      "loss": 2.2638,
+      "step": 260
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 0.9849622761444283,
+      "learning_rate": 1.9117183286805726e-05,
+      "loss": 2.148,
+      "step": 261
+    },
+    {
+      "epoch": 0.6230677764565993,
+      "grad_norm": 1.384730629468189,
+      "learning_rate": 1.9109296952954826e-05,
+      "loss": 2.2442,
+      "step": 262
+    },
+    {
+      "epoch": 0.6254458977407847,
+      "grad_norm": 1.066860211861379,
+      "learning_rate": 1.9101377192152033e-05,
+      "loss": 2.451,
+      "step": 263
+    },
+    {
+      "epoch": 0.6278240190249703,
+      "grad_norm": 1.1537333721148562,
+      "learning_rate": 1.909342403345925e-05,
+      "loss": 2.3076,
+      "step": 264
+    },
+    {
+      "epoch": 0.6302021403091558,
+      "grad_norm": 1.388798426336895,
+      "learning_rate": 1.9085437506060925e-05,
+      "loss": 2.3458,
+      "step": 265
+    },
+    {
+      "epoch": 0.6325802615933412,
+      "grad_norm": 1.057136103530956,
+      "learning_rate": 1.9077417639263966e-05,
+      "loss": 2.1901,
+      "step": 266
+    },
+    {
+      "epoch": 0.6349583828775267,
+      "grad_norm": 1.0267855180849,
+      "learning_rate": 1.906936446249761e-05,
+      "loss": 2.2597,
+      "step": 267
+    },
+    {
+      "epoch": 0.6373365041617123,
+      "grad_norm": 0.984822259782258,
+      "learning_rate": 1.906127800531333e-05,
+      "loss": 1.9602,
+      "step": 268
+    },
+    {
+      "epoch": 0.6397146254458977,
+      "grad_norm": 1.0914462759031294,
+      "learning_rate": 1.905315829738473e-05,
+      "loss": 2.4246,
+      "step": 269
+    },
+    {
+      "epoch": 0.6420927467300832,
+      "grad_norm": 1.007066333419377,
+      "learning_rate": 1.9045005368507418e-05,
+      "loss": 2.2638,
+      "step": 270
+    },
+    {
+      "epoch": 0.6444708680142688,
+      "grad_norm": 0.8517882887076221,
+      "learning_rate": 1.9036819248598914e-05,
+      "loss": 1.9905,
+      "step": 271
+    },
+    {
+      "epoch": 0.6468489892984542,
+      "grad_norm": 1.0217039586834744,
+      "learning_rate": 1.9028599967698533e-05,
+      "loss": 2.2177,
+      "step": 272
+    },
+    {
+      "epoch": 0.6492271105826397,
+      "grad_norm": 0.9930930260750681,
+      "learning_rate": 1.902034755596727e-05,
+      "loss": 2.093,
+      "step": 273
+    },
+    {
+      "epoch": 0.6516052318668252,
+      "grad_norm": 1.4476621256505835,
+      "learning_rate": 1.9012062043687713e-05,
+      "loss": 2.2535,
+      "step": 274
+    },
+    {
+      "epoch": 0.6539833531510107,
+      "grad_norm": 0.9679973366177126,
+      "learning_rate": 1.9003743461263887e-05,
+      "loss": 2.2304,
+      "step": 275
+    },
+    {
+      "epoch": 0.6563614744351962,
+      "grad_norm": 1.0576995279849084,
+      "learning_rate": 1.899539183922119e-05,
+      "loss": 2.0678,
+      "step": 276
+    },
+    {
+      "epoch": 0.6587395957193817,
+      "grad_norm": 0.9196543804308106,
+      "learning_rate": 1.8987007208206254e-05,
+      "loss": 2.1106,
+      "step": 277
+    },
+    {
+      "epoch": 0.6611177170035671,
+      "grad_norm": 1.0412924135468502,
+      "learning_rate": 1.8978589598986838e-05,
+      "loss": 2.3659,
+      "step": 278
+    },
+    {
+      "epoch": 0.6634958382877527,
+      "grad_norm": 1.1221322457101093,
+      "learning_rate": 1.8970139042451712e-05,
+      "loss": 2.0853,
+      "step": 279
+    },
+    {
+      "epoch": 0.6658739595719382,
+      "grad_norm": 0.8927332524290843,
+      "learning_rate": 1.8961655569610557e-05,
+      "loss": 2.2388,
+      "step": 280
+    },
+    {
+      "epoch": 0.6682520808561236,
+      "grad_norm": 1.0304293869359518,
+      "learning_rate": 1.8953139211593838e-05,
+      "loss": 2.381,
+      "step": 281
+    },
+    {
+      "epoch": 0.6706302021403091,
+      "grad_norm": 1.0508469543032843,
+      "learning_rate": 1.8944589999652687e-05,
+      "loss": 2.214,
+      "step": 282
+    },
+    {
+      "epoch": 0.6730083234244947,
+      "grad_norm": 0.8545452252368634,
+      "learning_rate": 1.8936007965158806e-05,
+      "loss": 2.0308,
+      "step": 283
+    },
+    {
+      "epoch": 0.6753864447086801,
+      "grad_norm": 1.3101079437708238,
+      "learning_rate": 1.8927393139604327e-05,
+      "loss": 2.3389,
+      "step": 284
+    },
+    {
+      "epoch": 0.6777645659928656,
+      "grad_norm": 0.8257605335721794,
+      "learning_rate": 1.8918745554601726e-05,
+      "loss": 1.9511,
+      "step": 285
+    },
+    {
+      "epoch": 0.6801426872770512,
+      "grad_norm": 0.9918391764529991,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 2.3694,
+      "step": 286
+    },
+    {
+      "epoch": 0.6825208085612366,
+      "grad_norm": 0.8841124075255721,
+      "learning_rate": 1.8901352233302964e-05,
+      "loss": 2.1977,
+      "step": 287
+    },
+    {
+      "epoch": 0.6848989298454221,
+      "grad_norm": 0.8100832981740426,
+      "learning_rate": 1.8892606560832335e-05,
+      "loss": 1.9161,
+      "step": 288
+    },
+    {
+      "epoch": 0.6872770511296076,
+      "grad_norm": 0.8738803349826288,
+      "learning_rate": 1.8883828256564413e-05,
+      "loss": 2.2419,
+      "step": 289
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 0.8961748350433363,
+      "learning_rate": 1.8875017352711547e-05,
+      "loss": 2.1639,
+      "step": 290
+    },
+    {
+      "epoch": 0.6920332936979786,
+      "grad_norm": 1.0902732587802257,
+      "learning_rate": 1.886617388160573e-05,
+      "loss": 2.4486,
+      "step": 291
+    },
+    {
+      "epoch": 0.6944114149821641,
+      "grad_norm": 0.9653296144522788,
+      "learning_rate": 1.8857297875698455e-05,
+      "loss": 2.0904,
+      "step": 292
+    },
+    {
+      "epoch": 0.6967895362663495,
+      "grad_norm": 1.0898489125890396,
+      "learning_rate": 1.8848389367560604e-05,
+      "loss": 2.2726,
+      "step": 293
+    },
+    {
+      "epoch": 0.6991676575505351,
+      "grad_norm": 0.8892982592369411,
+      "learning_rate": 1.883944838988232e-05,
+      "loss": 1.8982,
+      "step": 294
+    },
+    {
+      "epoch": 0.7015457788347206,
+      "grad_norm": 0.9481426246994946,
+      "learning_rate": 1.8830474975472904e-05,
+      "loss": 2.1466,
+      "step": 295
+    },
+    {
+      "epoch": 0.703923900118906,
+      "grad_norm": 0.8426682151678768,
+      "learning_rate": 1.8821469157260687e-05,
+      "loss": 2.1086,
+      "step": 296
+    },
+    {
+      "epoch": 0.7063020214030915,
+      "grad_norm": 1.0388273479050054,
+      "learning_rate": 1.8812430968292888e-05,
+      "loss": 2.3994,
+      "step": 297
+    },
+    {
+      "epoch": 0.7086801426872771,
+      "grad_norm": 0.9544054737799986,
+      "learning_rate": 1.8803360441735533e-05,
+      "loss": 2.0918,
+      "step": 298
+    },
+    {
+      "epoch": 0.7110582639714625,
+      "grad_norm": 0.9333443425947632,
+      "learning_rate": 1.8794257610873307e-05,
+      "loss": 2.2388,
+      "step": 299
+    },
+    {
+      "epoch": 0.713436385255648,
+      "grad_norm": 0.9411120276164652,
+      "learning_rate": 1.8785122509109425e-05,
+      "loss": 2.1787,
+      "step": 300
+    },
+    {
+      "epoch": 0.7158145065398336,
+      "grad_norm": 0.899942759501507,
+      "learning_rate": 1.877595516996554e-05,
+      "loss": 2.2007,
+      "step": 301
+    },
+    {
+      "epoch": 0.718192627824019,
+      "grad_norm": 0.952238040187914,
+      "learning_rate": 1.8766755627081586e-05,
+      "loss": 2.1491,
+      "step": 302
+    },
+    {
+      "epoch": 0.7205707491082045,
+      "grad_norm": 1.0704167928673332,
+      "learning_rate": 1.875752391421568e-05,
+      "loss": 2.4306,
+      "step": 303
+    },
+    {
+      "epoch": 0.72294887039239,
+      "grad_norm": 1.0126343580756334,
+      "learning_rate": 1.8748260065243985e-05,
+      "loss": 2.3258,
+      "step": 304
+    },
+    {
+      "epoch": 0.7253269916765755,
+      "grad_norm": 0.9203276329051738,
+      "learning_rate": 1.8738964114160586e-05,
+      "loss": 2.0469,
+      "step": 305
+    },
+    {
+      "epoch": 0.727705112960761,
+      "grad_norm": 1.0298534605384366,
+      "learning_rate": 1.8729636095077368e-05,
+      "loss": 2.3807,
+      "step": 306
+    },
+    {
+      "epoch": 0.7300832342449465,
+      "grad_norm": 0.9102516225979571,
+      "learning_rate": 1.8720276042223896e-05,
+      "loss": 2.0705,
+      "step": 307
+    },
+    {
+      "epoch": 0.7324613555291319,
+      "grad_norm": 0.8450936687346169,
+      "learning_rate": 1.8710883989947278e-05,
+      "loss": 1.9468,
+      "step": 308
+    },
+    {
+      "epoch": 0.7348394768133175,
+      "grad_norm": 10.37449210680128,
+      "learning_rate": 1.870145997271206e-05,
+      "loss": 2.2267,
+      "step": 309
+    },
+    {
+      "epoch": 0.737217598097503,
+      "grad_norm": 0.9380258905092027,
+      "learning_rate": 1.8692004025100054e-05,
+      "loss": 2.2396,
+      "step": 310
+    },
+    {
+      "epoch": 0.7395957193816884,
+      "grad_norm": 0.9949249727532327,
+      "learning_rate": 1.868251618181028e-05,
+      "loss": 2.2737,
+      "step": 311
+    },
+    {
+      "epoch": 0.7419738406658739,
+      "grad_norm": 0.8650417909396523,
+      "learning_rate": 1.8672996477658767e-05,
+      "loss": 2.2057,
+      "step": 312
+    },
+    {
+      "epoch": 0.7443519619500595,
+      "grad_norm": 0.9963224791105781,
+      "learning_rate": 1.866344494757848e-05,
+      "loss": 2.3958,
+      "step": 313
+    },
+    {
+      "epoch": 0.746730083234245,
+      "grad_norm": 0.9782743745256545,
+      "learning_rate": 1.8653861626619166e-05,
+      "loss": 2.105,
+      "step": 314
+    },
+    {
+      "epoch": 0.7491082045184304,
+      "grad_norm": 0.9991591823799537,
+      "learning_rate": 1.8644246549947226e-05,
+      "loss": 2.3339,
+      "step": 315
+    },
+    {
+      "epoch": 0.7491082045184304,
+      "eval_loss": 2.453089952468872,
+      "eval_runtime": 65.8189,
+      "eval_samples_per_second": 9.708,
+      "eval_steps_per_second": 1.215,
+      "step": 315
+    },
+    {
+      "epoch": 0.751486325802616,
+      "grad_norm": 0.8595275246052216,
+      "learning_rate": 1.8634599752845594e-05,
+      "loss": 2.2405,
+      "step": 316
+    },
+    {
+      "epoch": 0.7538644470868014,
+      "grad_norm": 0.9508091435584168,
+      "learning_rate": 1.86249212707136e-05,
+      "loss": 2.0259,
+      "step": 317
+    },
+    {
+      "epoch": 0.7562425683709869,
+      "grad_norm": 1.007143450486727,
+      "learning_rate": 1.861521113906684e-05,
+      "loss": 2.2581,
+      "step": 318
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 0.8694388672608973,
+      "learning_rate": 1.8605469393537062e-05,
+      "loss": 2.1028,
+      "step": 319
+    },
+    {
+      "epoch": 0.760998810939358,
+      "grad_norm": 1.1523113038633086,
+      "learning_rate": 1.8595696069872013e-05,
+      "loss": 2.5605,
+      "step": 320
+    },
+    {
+      "epoch": 0.7633769322235434,
+      "grad_norm": 1.0752431273595597,
+      "learning_rate": 1.8585891203935315e-05,
+      "loss": 2.179,
+      "step": 321
+    },
+    {
+      "epoch": 0.7657550535077289,
+      "grad_norm": 1.4087896659381576,
+      "learning_rate": 1.8576054831706348e-05,
+      "loss": 2.3727,
+      "step": 322
+    },
+    {
+      "epoch": 0.7681331747919143,
+      "grad_norm": 1.0235487680455349,
+      "learning_rate": 1.856618698928009e-05,
+      "loss": 2.0918,
+      "step": 323
+    },
+    {
+      "epoch": 0.7705112960760999,
+      "grad_norm": 0.9904332707712578,
+      "learning_rate": 1.8556287712867006e-05,
+      "loss": 2.2627,
+      "step": 324
+    },
+    {
+      "epoch": 0.7728894173602854,
+      "grad_norm": 0.9900875612198691,
+      "learning_rate": 1.8546357038792918e-05,
+      "loss": 2.2281,
+      "step": 325
+    },
+    {
+      "epoch": 0.7752675386444708,
+      "grad_norm": 0.8882946865776852,
+      "learning_rate": 1.8536395003498857e-05,
+      "loss": 2.2966,
+      "step": 326
+    },
+    {
+      "epoch": 0.7776456599286563,
+      "grad_norm": 0.9917259649244573,
+      "learning_rate": 1.8526401643540924e-05,
+      "loss": 2.2903,
+      "step": 327
+    },
+    {
+      "epoch": 0.7800237812128419,
+      "grad_norm": 1.0110335104954997,
+      "learning_rate": 1.8516376995590185e-05,
+      "loss": 2.2732,
+      "step": 328
+    },
+    {
+      "epoch": 0.7824019024970273,
+      "grad_norm": 1.0613748791083453,
+      "learning_rate": 1.8506321096432516e-05,
+      "loss": 2.1801,
+      "step": 329
+    },
+    {
+      "epoch": 0.7847800237812128,
+      "grad_norm": 0.8280074594268549,
+      "learning_rate": 1.849623398296846e-05,
+      "loss": 2.0413,
+      "step": 330
+    },
+    {
+      "epoch": 0.7871581450653984,
+      "grad_norm": 1.061771296332105,
+      "learning_rate": 1.84861156922131e-05,
+      "loss": 2.3187,
+      "step": 331
+    },
+    {
+      "epoch": 0.7895362663495838,
+      "grad_norm": 1.003700059201312,
+      "learning_rate": 1.8475966261295947e-05,
+      "loss": 2.2117,
+      "step": 332
+    },
+    {
+      "epoch": 0.7919143876337693,
+      "grad_norm": 1.0346094081558943,
+      "learning_rate": 1.8465785727460763e-05,
+      "loss": 2.35,
+      "step": 333
+    },
+    {
+      "epoch": 0.7942925089179548,
+      "grad_norm": 1.0628441168006169,
+      "learning_rate": 1.845557412806545e-05,
+      "loss": 2.2907,
+      "step": 334
+    },
+    {
+      "epoch": 0.7966706302021404,
+      "grad_norm": 0.9035239279839531,
+      "learning_rate": 1.8445331500581905e-05,
+      "loss": 2.3114,
+      "step": 335
+    },
+    {
+      "epoch": 0.7990487514863258,
+      "grad_norm": 0.8520631758698989,
+      "learning_rate": 1.8435057882595885e-05,
+      "loss": 2.236,
+      "step": 336
+    },
+    {
+      "epoch": 0.8014268727705113,
+      "grad_norm": 0.9909579691053704,
+      "learning_rate": 1.8424753311806867e-05,
+      "loss": 2.164,
+      "step": 337
+    },
+    {
+      "epoch": 0.8038049940546967,
+      "grad_norm": 1.3497517322085937,
+      "learning_rate": 1.8414417826027907e-05,
+      "loss": 2.1371,
+      "step": 338
+    },
+    {
+      "epoch": 0.8061831153388823,
+      "grad_norm": 0.9733174734597209,
+      "learning_rate": 1.840405146318552e-05,
+      "loss": 2.2605,
+      "step": 339
+    },
+    {
+      "epoch": 0.8085612366230678,
+      "grad_norm": 0.991877145495908,
+      "learning_rate": 1.8393654261319504e-05,
+      "loss": 2.2597,
+      "step": 340
+    },
+    {
+      "epoch": 0.8109393579072532,
+      "grad_norm": 0.8567918308119495,
+      "learning_rate": 1.8383226258582833e-05,
+      "loss": 2.0034,
+      "step": 341
+    },
+    {
+      "epoch": 0.8133174791914387,
+      "grad_norm": 1.065116900935523,
+      "learning_rate": 1.837276749324151e-05,
+      "loss": 2.3297,
+      "step": 342
+    },
+    {
+      "epoch": 0.8156956004756243,
+      "grad_norm": 0.945929516741479,
+      "learning_rate": 1.8362278003674417e-05,
+      "loss": 2.1654,
+      "step": 343
+    },
+    {
+      "epoch": 0.8180737217598097,
+      "grad_norm": 1.2321854339549771,
+      "learning_rate": 1.8351757828373183e-05,
+      "loss": 2.3218,
+      "step": 344
+    },
+    {
+      "epoch": 0.8204518430439952,
+      "grad_norm": 0.8769538008656999,
+      "learning_rate": 1.8341207005942033e-05,
+      "loss": 1.9733,
+      "step": 345
+    },
+    {
+      "epoch": 0.8228299643281808,
+      "grad_norm": 0.9563552488015744,
+      "learning_rate": 1.8330625575097663e-05,
+      "loss": 1.8932,
+      "step": 346
+    },
+    {
+      "epoch": 0.8252080856123662,
+      "grad_norm": 0.9405864126962645,
+      "learning_rate": 1.8320013574669083e-05,
+      "loss": 2.2498,
+      "step": 347
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 0.910096309278826,
+      "learning_rate": 1.8309371043597472e-05,
+      "loss": 2.0247,
+      "step": 348
+    },
+    {
+      "epoch": 0.8299643281807372,
+      "grad_norm": 0.8784633584470368,
+      "learning_rate": 1.829869802093606e-05,
+      "loss": 1.9894,
+      "step": 349
+    },
+    {
+      "epoch": 0.8323424494649228,
+      "grad_norm": 0.8541120126588495,
+      "learning_rate": 1.8287994545849948e-05,
+      "loss": 2.0511,
+      "step": 350
+    },
+    {
+      "epoch": 0.8347205707491082,
+      "grad_norm": 0.9979547733824415,
+      "learning_rate": 1.8277260657615993e-05,
+      "loss": 2.0893,
+      "step": 351
+    },
+    {
+      "epoch": 0.8370986920332937,
+      "grad_norm": 1.0836986822911394,
+      "learning_rate": 1.826649639562266e-05,
+      "loss": 2.3076,
+      "step": 352
+    },
+    {
+      "epoch": 0.8394768133174791,
+      "grad_norm": 1.035582585906488,
+      "learning_rate": 1.825570179936986e-05,
+      "loss": 2.2928,
+      "step": 353
+    },
+    {
+      "epoch": 0.8418549346016647,
+      "grad_norm": 0.8458996980228227,
+      "learning_rate": 1.8244876908468826e-05,
+      "loss": 2.0421,
+      "step": 354
+    },
+    {
+      "epoch": 0.8442330558858502,
+      "grad_norm": 0.9377520676275963,
+      "learning_rate": 1.8234021762641946e-05,
+      "loss": 2.2872,
+      "step": 355
+    },
+    {
+      "epoch": 0.8466111771700356,
+      "grad_norm": 0.9662487818099018,
+      "learning_rate": 1.8223136401722648e-05,
+      "loss": 2.2357,
+      "step": 356
+    },
+    {
+      "epoch": 0.8489892984542212,
+      "grad_norm": 0.9373837869014561,
+      "learning_rate": 1.8212220865655224e-05,
+      "loss": 2.3807,
+      "step": 357
+    },
+    {
+      "epoch": 0.8513674197384067,
+      "grad_norm": 0.8767072116092074,
+      "learning_rate": 1.8201275194494695e-05,
+      "loss": 2.0453,
+      "step": 358
+    },
+    {
+      "epoch": 0.8537455410225921,
+      "grad_norm": 0.9164578385360742,
+      "learning_rate": 1.8190299428406667e-05,
+      "loss": 2.1177,
+      "step": 359
+    },
+    {
+      "epoch": 0.8561236623067776,
+      "grad_norm": 2.207352437074081,
+      "learning_rate": 1.8179293607667177e-05,
+      "loss": 2.2001,
+      "step": 360
+    },
+    {
+      "epoch": 0.8585017835909632,
+      "grad_norm": 0.983155025111798,
+      "learning_rate": 1.8168257772662556e-05,
+      "loss": 2.3009,
+      "step": 361
+    },
+    {
+      "epoch": 0.8608799048751486,
+      "grad_norm": 0.9734880504382539,
+      "learning_rate": 1.8157191963889265e-05,
+      "loss": 2.3093,
+      "step": 362
+    },
+    {
+      "epoch": 0.8632580261593341,
+      "grad_norm": 0.8887247626382502,
+      "learning_rate": 1.8146096221953767e-05,
+      "loss": 2.1673,
+      "step": 363
+    },
+    {
+      "epoch": 0.8656361474435196,
+      "grad_norm": 0.9468700496948975,
+      "learning_rate": 1.8134970587572345e-05,
+      "loss": 2.2193,
+      "step": 364
+    },
+    {
+      "epoch": 0.8680142687277052,
+      "grad_norm": 0.9774148680996411,
+      "learning_rate": 1.8123815101570996e-05,
+      "loss": 2.2185,
+      "step": 365
+    },
+    {
+      "epoch": 0.8703923900118906,
+      "grad_norm": 1.0377594097114105,
+      "learning_rate": 1.8112629804885248e-05,
+      "loss": 2.1385,
+      "step": 366
+    },
+    {
+      "epoch": 0.8727705112960761,
+      "grad_norm": 0.9257473414011718,
+      "learning_rate": 1.8101414738560018e-05,
+      "loss": 2.309,
+      "step": 367
+    },
+    {
+      "epoch": 0.8751486325802615,
+      "grad_norm": 0.8751605326561603,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 2.0981,
+      "step": 368
+    },
+    {
+      "epoch": 0.8775267538644471,
+      "grad_norm": 0.8318450854470809,
+      "learning_rate": 1.8078895461716867e-05,
+      "loss": 2.0499,
+      "step": 369
+    },
+    {
+      "epoch": 0.8799048751486326,
+      "grad_norm": 0.9229990617872996,
+      "learning_rate": 1.8067591333834382e-05,
+      "loss": 2.0931,
+      "step": 370
+    },
+    {
+      "epoch": 0.882282996432818,
+      "grad_norm": 0.8885903764744792,
+      "learning_rate": 1.8056257601583004e-05,
+      "loss": 2.1094,
+      "step": 371
+    },
+    {
+      "epoch": 0.8846611177170036,
+      "grad_norm": 0.8862767802235494,
+      "learning_rate": 1.8044894306552338e-05,
+      "loss": 2.1633,
+      "step": 372
+    },
+    {
+      "epoch": 0.8870392390011891,
+      "grad_norm": 0.8644263027922633,
+      "learning_rate": 1.8033501490440478e-05,
+      "loss": 2.1869,
+      "step": 373
+    },
+    {
+      "epoch": 0.8894173602853745,
+      "grad_norm": 0.8908757391357947,
+      "learning_rate": 1.802207919505385e-05,
+      "loss": 2.0867,
+      "step": 374
+    },
+    {
+      "epoch": 0.89179548156956,
+      "grad_norm": 0.9706439808339724,
+      "learning_rate": 1.801062746230705e-05,
+      "loss": 2.2817,
+      "step": 375
+    },
+    {
+      "epoch": 0.8941736028537456,
+      "grad_norm": 0.9665648406193532,
+      "learning_rate": 1.79991463342227e-05,
+      "loss": 2.109,
+      "step": 376
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 0.8628669010410803,
+      "learning_rate": 1.798763585293128e-05,
+      "loss": 2.0049,
+      "step": 377
+    },
+    {
+      "epoch": 0.8989298454221165,
+      "grad_norm": 0.9905082003227874,
+      "learning_rate": 1.7976096060671e-05,
+      "loss": 2.3004,
+      "step": 378
+    },
+    {
+      "epoch": 0.901307966706302,
+      "grad_norm": 0.9857769539636477,
+      "learning_rate": 1.7964526999787606e-05,
+      "loss": 2.2076,
+      "step": 379
+    },
+    {
+      "epoch": 0.9036860879904876,
+      "grad_norm": 0.9874365687027726,
+      "learning_rate": 1.7952928712734266e-05,
+      "loss": 2.2918,
+      "step": 380
+    },
+    {
+      "epoch": 0.906064209274673,
+      "grad_norm": 0.8854059768194814,
+      "learning_rate": 1.7941301242071384e-05,
+      "loss": 2.1416,
+      "step": 381
+    },
+    {
+      "epoch": 0.9084423305588585,
+      "grad_norm": 0.8891010909161526,
+      "learning_rate": 1.792964463046646e-05,
+      "loss": 2.2335,
+      "step": 382
+    },
+    {
+      "epoch": 0.9108204518430439,
+      "grad_norm": 0.8836139495461457,
+      "learning_rate": 1.7917958920693923e-05,
+      "loss": 2.0156,
+      "step": 383
+    },
+    {
+      "epoch": 0.9131985731272295,
+      "grad_norm": 0.862309723766035,
+      "learning_rate": 1.790624415563498e-05,
+      "loss": 2.2668,
+      "step": 384
+    },
+    {
+      "epoch": 0.915576694411415,
+      "grad_norm": 0.8736380448815635,
+      "learning_rate": 1.7894500378277463e-05,
+      "loss": 2.1338,
+      "step": 385
+    },
+    {
+      "epoch": 0.9179548156956004,
+      "grad_norm": 0.8450572821559386,
+      "learning_rate": 1.7882727631715655e-05,
+      "loss": 2.0142,
+      "step": 386
+    },
+    {
+      "epoch": 0.920332936979786,
+      "grad_norm": 0.9862096866699867,
+      "learning_rate": 1.7870925959150155e-05,
+      "loss": 1.9915,
+      "step": 387
+    },
+    {
+      "epoch": 0.9227110582639715,
+      "grad_norm": 1.0092700839816597,
+      "learning_rate": 1.7859095403887697e-05,
+      "loss": 2.0294,
+      "step": 388
+    },
+    {
+      "epoch": 0.925089179548157,
+      "grad_norm": 1.1074789492464077,
+      "learning_rate": 1.7847236009341007e-05,
+      "loss": 2.2817,
+      "step": 389
+    },
+    {
+      "epoch": 0.9274673008323424,
+      "grad_norm": 1.1905201839698938,
+      "learning_rate": 1.7835347819028642e-05,
+      "loss": 2.1793,
+      "step": 390
+    },
+    {
+      "epoch": 0.929845422116528,
+      "grad_norm": 0.9783879172577963,
+      "learning_rate": 1.7823430876574815e-05,
+      "loss": 2.0936,
+      "step": 391
+    },
+    {
+      "epoch": 0.9322235434007135,
+      "grad_norm": 0.8747590280954227,
+      "learning_rate": 1.7811485225709255e-05,
+      "loss": 2.0459,
+      "step": 392
+    },
+    {
+      "epoch": 0.9346016646848989,
+      "grad_norm": 0.9379273246769285,
+      "learning_rate": 1.7799510910267032e-05,
+      "loss": 2.3024,
+      "step": 393
+    },
+    {
+      "epoch": 0.9369797859690844,
+      "grad_norm": 1.0154785871015584,
+      "learning_rate": 1.778750797418841e-05,
+      "loss": 2.2336,
+      "step": 394
+    },
+    {
+      "epoch": 0.93935790725327,
+      "grad_norm": 2.334020052582292,
+      "learning_rate": 1.7775476461518668e-05,
+      "loss": 2.2146,
+      "step": 395
+    },
+    {
+      "epoch": 0.9417360285374554,
+      "grad_norm": 1.028424946167679,
+      "learning_rate": 1.7763416416407953e-05,
+      "loss": 2.148,
+      "step": 396
+    },
+    {
+      "epoch": 0.9441141498216409,
+      "grad_norm": 0.9939249719466411,
+      "learning_rate": 1.7751327883111117e-05,
+      "loss": 2.3384,
+      "step": 397
+    },
+    {
+      "epoch": 0.9464922711058263,
+      "grad_norm": 0.9622526498121815,
+      "learning_rate": 1.773921090598754e-05,
+      "loss": 2.2333,
+      "step": 398
+    },
+    {
+      "epoch": 0.9488703923900119,
+      "grad_norm": 0.9627114271614428,
+      "learning_rate": 1.7727065529500986e-05,
+      "loss": 2.0078,
+      "step": 399
+    },
+    {
+      "epoch": 0.9512485136741974,
+      "grad_norm": 1.0739248368345071,
+      "learning_rate": 1.7714891798219432e-05,
+      "loss": 2.431,
+      "step": 400
+    },
+    {
+      "epoch": 0.9536266349583828,
+      "grad_norm": 0.9459448330629742,
+      "learning_rate": 1.7702689756814898e-05,
+      "loss": 2.195,
+      "step": 401
+    },
+    {
+      "epoch": 0.9560047562425684,
+      "grad_norm": 0.985098727152756,
+      "learning_rate": 1.7690459450063297e-05,
+      "loss": 2.378,
+      "step": 402
+    },
+    {
+      "epoch": 0.9583828775267539,
+      "grad_norm": 0.9004480308740594,
+      "learning_rate": 1.7678200922844256e-05,
+      "loss": 2.0375,
+      "step": 403
+    },
+    {
+      "epoch": 0.9607609988109393,
+      "grad_norm": 1.0816635115771127,
+      "learning_rate": 1.7665914220140964e-05,
+      "loss": 2.2932,
+      "step": 404
+    },
+    {
+      "epoch": 0.9631391200951248,
+      "grad_norm": 0.856817467594963,
+      "learning_rate": 1.7653599387039993e-05,
+      "loss": 2.1395,
+      "step": 405
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 0.9816466176158317,
+      "learning_rate": 1.764125646873115e-05,
+      "loss": 2.2032,
+      "step": 406
+    },
+    {
+      "epoch": 0.9678953626634959,
+      "grad_norm": 1.0159180902253635,
+      "learning_rate": 1.7628885510507295e-05,
+      "loss": 2.3814,
+      "step": 407
+    },
+    {
+      "epoch": 0.9702734839476813,
+      "grad_norm": 0.8765185622300491,
+      "learning_rate": 1.7616486557764187e-05,
+      "loss": 2.1377,
+      "step": 408
+    },
+    {
+      "epoch": 0.9726516052318668,
+      "grad_norm": 0.8581340453429611,
+      "learning_rate": 1.7604059656000313e-05,
+      "loss": 2.1687,
+      "step": 409
+    },
+    {
+      "epoch": 0.9750297265160524,
+      "grad_norm": 1.0106580156311586,
+      "learning_rate": 1.7591604850816705e-05,
+      "loss": 2.4389,
+      "step": 410
+    },
+    {
+      "epoch": 0.9774078478002378,
+      "grad_norm": 0.9102932420528898,
+      "learning_rate": 1.757912218791681e-05,
+      "loss": 2.1774,
+      "step": 411
+    },
+    {
+      "epoch": 0.9797859690844233,
+      "grad_norm": 0.8897815759878576,
+      "learning_rate": 1.7566611713106287e-05,
+      "loss": 2.0938,
+      "step": 412
+    },
+    {
+      "epoch": 0.9821640903686087,
+      "grad_norm": 0.8151179975871685,
+      "learning_rate": 1.7554073472292854e-05,
+      "loss": 2.0923,
+      "step": 413
+    },
+    {
+      "epoch": 0.9845422116527943,
+      "grad_norm": 0.9325421154931588,
+      "learning_rate": 1.7541507511486114e-05,
+      "loss": 2.2798,
+      "step": 414
+    },
+    {
+      "epoch": 0.9869203329369798,
+      "grad_norm": 0.9169184208068022,
+      "learning_rate": 1.75289138767974e-05,
+      "loss": 2.2266,
+      "step": 415
+    },
+    {
+      "epoch": 0.9892984542211652,
+      "grad_norm": 0.9539647094698775,
+      "learning_rate": 1.7516292614439586e-05,
+      "loss": 2.3459,
+      "step": 416
+    },
+    {
+      "epoch": 0.9916765755053508,
+      "grad_norm": 1.0372921455846975,
+      "learning_rate": 1.7503643770726924e-05,
+      "loss": 2.2152,
+      "step": 417
+    },
+    {
+      "epoch": 0.9940546967895363,
+      "grad_norm": 0.8903926236033542,
+      "learning_rate": 1.7490967392074897e-05,
+      "loss": 2.1959,
+      "step": 418
+    },
+    {
+      "epoch": 0.9964328180737217,
+      "grad_norm": 0.8491943369235393,
+      "learning_rate": 1.7478263525000003e-05,
+      "loss": 2.265,
+      "step": 419
+    },
+    {
+      "epoch": 0.9988109393579072,
+      "grad_norm": 0.9138223778798192,
+      "learning_rate": 1.7465532216119628e-05,
+      "loss": 2.2167,
+      "step": 420
+    },
+    {
+      "epoch": 0.9988109393579072,
+      "eval_loss": 2.4223339557647705,
+      "eval_runtime": 65.3623,
+      "eval_samples_per_second": 9.776,
+      "eval_steps_per_second": 1.224,
+      "step": 420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9138223778798192,
+      "learning_rate": 1.7452773512151847e-05,
+      "loss": 2.272,
+      "step": 421
+    },
+    {
+      "epoch": 1.0023781212841856,
+      "grad_norm": 1.628296924382809,
+      "learning_rate": 1.7439987459915265e-05,
+      "loss": 1.7689,
+      "step": 422
+    },
+    {
+      "epoch": 1.004756242568371,
+      "grad_norm": 1.0685556763602093,
+      "learning_rate": 1.7427174106328846e-05,
+      "loss": 1.952,
+      "step": 423
+    },
+    {
+      "epoch": 1.0071343638525565,
+      "grad_norm": 1.1222612546095956,
+      "learning_rate": 1.7414333498411734e-05,
+      "loss": 2.0925,
+      "step": 424
+    },
+    {
+      "epoch": 1.009512485136742,
+      "grad_norm": 1.0714061881839094,
+      "learning_rate": 1.7401465683283085e-05,
+      "loss": 1.9147,
+      "step": 425
+    },
+    {
+      "epoch": 1.0118906064209274,
+      "grad_norm": 1.7138310113119204,
+      "learning_rate": 1.7388570708161895e-05,
+      "loss": 2.0451,
+      "step": 426
+    },
+    {
+      "epoch": 1.014268727705113,
+      "grad_norm": 1.2402516993910184,
+      "learning_rate": 1.7375648620366817e-05,
+      "loss": 2.0889,
+      "step": 427
+    },
+    {
+      "epoch": 1.0166468489892984,
+      "grad_norm": 1.120370001804281,
+      "learning_rate": 1.7362699467316005e-05,
+      "loss": 2.0449,
+      "step": 428
+    },
+    {
+      "epoch": 1.019024970273484,
+      "grad_norm": 1.1942571073271016,
+      "learning_rate": 1.7349723296526923e-05,
+      "loss": 1.9913,
+      "step": 429
+    },
+    {
+      "epoch": 1.0214030915576695,
+      "grad_norm": 1.193074546908054,
+      "learning_rate": 1.7336720155616186e-05,
+      "loss": 2.0203,
+      "step": 430
+    },
+    {
+      "epoch": 1.0237812128418549,
+      "grad_norm": 0.8336179192625718,
+      "learning_rate": 1.732369009229937e-05,
+      "loss": 1.9839,
+      "step": 431
+    },
+    {
+      "epoch": 1.0261593341260404,
+      "grad_norm": 1.0440055245119049,
+      "learning_rate": 1.731063315439084e-05,
+      "loss": 1.9837,
+      "step": 432
+    },
+    {
+      "epoch": 1.028537455410226,
+      "grad_norm": 1.0017223079381312,
+      "learning_rate": 1.729754938980359e-05,
+      "loss": 2.0407,
+      "step": 433
+    },
+    {
+      "epoch": 1.0309155766944114,
+      "grad_norm": 1.0631773992935534,
+      "learning_rate": 1.7284438846549045e-05,
+      "loss": 2.0097,
+      "step": 434
+    },
+    {
+      "epoch": 1.033293697978597,
+      "grad_norm": 0.9980542007217309,
+      "learning_rate": 1.7271301572736904e-05,
+      "loss": 2.0821,
+      "step": 435
+    },
+    {
+      "epoch": 1.0356718192627823,
+      "grad_norm": 1.176402924330386,
+      "learning_rate": 1.725813761657495e-05,
+      "loss": 1.8594,
+      "step": 436
+    },
+    {
+      "epoch": 1.0380499405469679,
+      "grad_norm": 0.9752892126553369,
+      "learning_rate": 1.7244947026368878e-05,
+      "loss": 1.9673,
+      "step": 437
+    },
+    {
+      "epoch": 1.0404280618311534,
+      "grad_norm": 0.8689953910803522,
+      "learning_rate": 1.723172985052212e-05,
+      "loss": 1.9404,
+      "step": 438
+    },
+    {
+      "epoch": 1.0428061831153388,
+      "grad_norm": 1.2475813437261176,
+      "learning_rate": 1.7218486137535663e-05,
+      "loss": 2.0084,
+      "step": 439
+    },
+    {
+      "epoch": 1.0451843043995244,
+      "grad_norm": 1.0792429756941824,
+      "learning_rate": 1.720521593600787e-05,
+      "loss": 1.7713,
+      "step": 440
+    },
+    {
+      "epoch": 1.04756242568371,
+      "grad_norm": 0.9766534238863007,
+      "learning_rate": 1.7191919294634318e-05,
+      "loss": 1.9303,
+      "step": 441
+    },
+    {
+      "epoch": 1.0499405469678953,
+      "grad_norm": 1.0433496430768199,
+      "learning_rate": 1.717859626220759e-05,
+      "loss": 2.0188,
+      "step": 442
+    },
+    {
+      "epoch": 1.0523186682520809,
+      "grad_norm": 1.119386810190894,
+      "learning_rate": 1.7165246887617124e-05,
+      "loss": 1.9778,
+      "step": 443
+    },
+    {
+      "epoch": 1.0546967895362664,
+      "grad_norm": 0.9743730355346135,
+      "learning_rate": 1.715187121984901e-05,
+      "loss": 1.8941,
+      "step": 444
+    },
+    {
+      "epoch": 1.0570749108204518,
+      "grad_norm": 1.0796463722814744,
+      "learning_rate": 1.7138469307985832e-05,
+      "loss": 1.9573,
+      "step": 445
+    },
+    {
+      "epoch": 1.0594530321046374,
+      "grad_norm": 1.0491377417614562,
+      "learning_rate": 1.7125041201206478e-05,
+      "loss": 1.8715,
+      "step": 446
+    },
+    {
+      "epoch": 1.0618311533888227,
+      "grad_norm": 1.0722587649757946,
+      "learning_rate": 1.711158694878595e-05,
+      "loss": 1.892,
+      "step": 447
+    },
+    {
+      "epoch": 1.0642092746730083,
+      "grad_norm": 0.9251557954985203,
+      "learning_rate": 1.7098106600095204e-05,
+      "loss": 1.8922,
+      "step": 448
+    },
+    {
+      "epoch": 1.0665873959571939,
+      "grad_norm": 1.0363952883035685,
+      "learning_rate": 1.708460020460095e-05,
+      "loss": 1.9076,
+      "step": 449
+    },
+    {
+      "epoch": 1.0689655172413792,
+      "grad_norm": 1.0688439486649002,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.9799,
+      "step": 450
+    },
+    {
+      "epoch": 1.0713436385255648,
+      "grad_norm": 1.1027957836331184,
+      "learning_rate": 1.7057509471546476e-05,
+      "loss": 1.9104,
+      "step": 451
+    },
+    {
+      "epoch": 1.0737217598097504,
+      "grad_norm": 1.618486595162837,
+      "learning_rate": 1.7043925233396855e-05,
+      "loss": 2.0171,
+      "step": 452
+    },
+    {
+      "epoch": 1.0760998810939357,
+      "grad_norm": 1.0017778693822061,
+      "learning_rate": 1.703031514726455e-05,
+      "loss": 2.0216,
+      "step": 453
+    },
+    {
+      "epoch": 1.0784780023781213,
+      "grad_norm": 1.0070735770943433,
+      "learning_rate": 1.7016679263092352e-05,
+      "loss": 2.0046,
+      "step": 454
+    },
+    {
+      "epoch": 1.0808561236623069,
+      "grad_norm": 1.1684678408376339,
+      "learning_rate": 1.700301763091771e-05,
+      "loss": 2.0023,
+      "step": 455
+    },
+    {
+      "epoch": 1.0832342449464922,
+      "grad_norm": 1.0525233453838523,
+      "learning_rate": 1.6989330300872576e-05,
+      "loss": 1.7791,
+      "step": 456
+    },
+    {
+      "epoch": 1.0856123662306778,
+      "grad_norm": 0.9255203454055275,
+      "learning_rate": 1.6975617323183175e-05,
+      "loss": 1.9353,
+      "step": 457
+    },
+    {
+      "epoch": 1.0879904875148632,
+      "grad_norm": 1.0726155742434451,
+      "learning_rate": 1.6961878748169868e-05,
+      "loss": 1.8358,
+      "step": 458
+    },
+    {
+      "epoch": 1.0903686087990487,
+      "grad_norm": 1.0856909362569107,
+      "learning_rate": 1.694811462624694e-05,
+      "loss": 2.0117,
+      "step": 459
+    },
+    {
+      "epoch": 1.0927467300832343,
+      "grad_norm": 1.0205381713143487,
+      "learning_rate": 1.6934325007922418e-05,
+      "loss": 1.9674,
+      "step": 460
+    },
+    {
+      "epoch": 1.0951248513674197,
+      "grad_norm": 1.2186466927066344,
+      "learning_rate": 1.6920509943797893e-05,
+      "loss": 2.0536,
+      "step": 461
+    },
+    {
+      "epoch": 1.0975029726516052,
+      "grad_norm": 1.0835683944583578,
+      "learning_rate": 1.690666948456833e-05,
+      "loss": 2.0184,
+      "step": 462
+    },
+    {
+      "epoch": 1.0998810939357908,
+      "grad_norm": 0.9540266090666372,
+      "learning_rate": 1.689280368102188e-05,
+      "loss": 1.7483,
+      "step": 463
+    },
+    {
+      "epoch": 1.1022592152199762,
+      "grad_norm": 1.1931606541151196,
+      "learning_rate": 1.6878912584039698e-05,
+      "loss": 1.8391,
+      "step": 464
+    },
+    {
+      "epoch": 1.1046373365041617,
+      "grad_norm": 0.9898663043482883,
+      "learning_rate": 1.6864996244595757e-05,
+      "loss": 2.0927,
+      "step": 465
+    },
+    {
+      "epoch": 1.1070154577883473,
+      "grad_norm": 1.1388354501174491,
+      "learning_rate": 1.6851054713756653e-05,
+      "loss": 2.1605,
+      "step": 466
+    },
+    {
+      "epoch": 1.1093935790725327,
+      "grad_norm": 1.0945133712632522,
+      "learning_rate": 1.6837088042681423e-05,
+      "loss": 1.9853,
+      "step": 467
+    },
+    {
+      "epoch": 1.1117717003567182,
+      "grad_norm": 1.1984401216196918,
+      "learning_rate": 1.6823096282621366e-05,
+      "loss": 2.0257,
+      "step": 468
+    },
+    {
+      "epoch": 1.1141498216409036,
+      "grad_norm": 0.8917099119377462,
+      "learning_rate": 1.6809079484919834e-05,
+      "loss": 1.8659,
+      "step": 469
+    },
+    {
+      "epoch": 1.1165279429250892,
+      "grad_norm": 1.0116433318578837,
+      "learning_rate": 1.679503770101206e-05,
+      "loss": 2.2598,
+      "step": 470
+    },
+    {
+      "epoch": 1.1189060642092747,
+      "grad_norm": 3.175396684459374,
+      "learning_rate": 1.6780970982424966e-05,
+      "loss": 1.9356,
+      "step": 471
+    },
+    {
+      "epoch": 1.12128418549346,
+      "grad_norm": 1.2346613722854758,
+      "learning_rate": 1.6766879380776983e-05,
+      "loss": 1.8155,
+      "step": 472
+    },
+    {
+      "epoch": 1.1236623067776457,
+      "grad_norm": 1.006693147566446,
+      "learning_rate": 1.675276294777783e-05,
+      "loss": 1.8518,
+      "step": 473
+    },
+    {
+      "epoch": 1.1260404280618312,
+      "grad_norm": 0.8894722137287147,
+      "learning_rate": 1.6738621735228363e-05,
+      "loss": 2.0505,
+      "step": 474
+    },
+    {
+      "epoch": 1.1284185493460166,
+      "grad_norm": 1.103577294552584,
+      "learning_rate": 1.672445579502036e-05,
+      "loss": 1.9912,
+      "step": 475
+    },
+    {
+      "epoch": 1.1307966706302022,
+      "grad_norm": 1.096718928499225,
+      "learning_rate": 1.671026517913634e-05,
+      "loss": 1.8938,
+      "step": 476
+    },
+    {
+      "epoch": 1.1331747919143877,
+      "grad_norm": 0.9817288455594879,
+      "learning_rate": 1.669604993964937e-05,
+      "loss": 1.8101,
+      "step": 477
+    },
+    {
+      "epoch": 1.135552913198573,
+      "grad_norm": 1.1460045828685286,
+      "learning_rate": 1.668181012872288e-05,
+      "loss": 2.2724,
+      "step": 478
+    },
+    {
+      "epoch": 1.1379310344827587,
+      "grad_norm": 1.1002427752644104,
+      "learning_rate": 1.666754579861045e-05,
+      "loss": 2.0842,
+      "step": 479
+    },
+    {
+      "epoch": 1.140309155766944,
+      "grad_norm": 1.0583766063889544,
+      "learning_rate": 1.6653257001655652e-05,
+      "loss": 2.006,
+      "step": 480
+    },
+    {
+      "epoch": 1.1426872770511296,
+      "grad_norm": 1.0862329799584287,
+      "learning_rate": 1.6638943790291838e-05,
+      "loss": 2.074,
+      "step": 481
+    },
+    {
+      "epoch": 1.1450653983353152,
+      "grad_norm": 1.215641711818309,
+      "learning_rate": 1.6624606217041933e-05,
+      "loss": 1.9628,
+      "step": 482
+    },
+    {
+      "epoch": 1.1474435196195005,
+      "grad_norm": 0.9889402798664048,
+      "learning_rate": 1.661024433451828e-05,
+      "loss": 2.0147,
+      "step": 483
+    },
+    {
+      "epoch": 1.149821640903686,
+      "grad_norm": 0.9856498299194657,
+      "learning_rate": 1.6595858195422414e-05,
+      "loss": 1.8529,
+      "step": 484
+    },
+    {
+      "epoch": 1.1521997621878715,
+      "grad_norm": 1.0678057323434258,
+      "learning_rate": 1.6581447852544877e-05,
+      "loss": 1.7709,
+      "step": 485
+    },
+    {
+      "epoch": 1.154577883472057,
+      "grad_norm": 0.8386080848257256,
+      "learning_rate": 1.6567013358765045e-05,
+      "loss": 1.9122,
+      "step": 486
+    },
+    {
+      "epoch": 1.1569560047562426,
+      "grad_norm": 1.0694865554443245,
+      "learning_rate": 1.6552554767050898e-05,
+      "loss": 1.8617,
+      "step": 487
+    },
+    {
+      "epoch": 1.1593341260404282,
+      "grad_norm": 1.9413752963807895,
+      "learning_rate": 1.6538072130458853e-05,
+      "loss": 1.9043,
+      "step": 488
+    },
+    {
+      "epoch": 1.1617122473246135,
+      "grad_norm": 1.04735712587821,
+      "learning_rate": 1.6523565502133562e-05,
+      "loss": 1.8987,
+      "step": 489
+    },
+    {
+      "epoch": 1.164090368608799,
+      "grad_norm": 0.9832837853313156,
+      "learning_rate": 1.6509034935307716e-05,
+      "loss": 2.1317,
+      "step": 490
+    },
+    {
+      "epoch": 1.1664684898929845,
+      "grad_norm": 1.1810675679989917,
+      "learning_rate": 1.6494480483301836e-05,
+      "loss": 1.9238,
+      "step": 491
+    },
+    {
+      "epoch": 1.16884661117717,
+      "grad_norm": 0.9165494247402323,
+      "learning_rate": 1.6479902199524116e-05,
+      "loss": 2.0648,
+      "step": 492
+    },
+    {
+      "epoch": 1.1712247324613556,
+      "grad_norm": 0.9636466082435201,
+      "learning_rate": 1.6465300137470178e-05,
+      "loss": 2.0287,
+      "step": 493
+    },
+    {
+      "epoch": 1.173602853745541,
+      "grad_norm": 0.9807575958399394,
+      "learning_rate": 1.645067435072291e-05,
+      "loss": 1.8887,
+      "step": 494
+    },
+    {
+      "epoch": 1.1759809750297265,
+      "grad_norm": 0.9586275603088916,
+      "learning_rate": 1.6436024892952256e-05,
+      "loss": 1.8363,
+      "step": 495
+    },
+    {
+      "epoch": 1.178359096313912,
+      "grad_norm": 0.9457623737642574,
+      "learning_rate": 1.6421351817915025e-05,
+      "loss": 2.146,
+      "step": 496
+    },
+    {
+      "epoch": 1.1807372175980975,
+      "grad_norm": 1.0017147042513856,
+      "learning_rate": 1.6406655179454694e-05,
+      "loss": 2.0542,
+      "step": 497
+    },
+    {
+      "epoch": 1.183115338882283,
+      "grad_norm": 1.9322307606799163,
+      "learning_rate": 1.6391935031501193e-05,
+      "loss": 1.9879,
+      "step": 498
+    },
+    {
+      "epoch": 1.1854934601664684,
+      "grad_norm": 1.084335177141969,
+      "learning_rate": 1.6377191428070734e-05,
+      "loss": 2.0715,
+      "step": 499
+    },
+    {
+      "epoch": 1.187871581450654,
+      "grad_norm": 0.9850991570942521,
+      "learning_rate": 1.63624244232656e-05,
+      "loss": 1.9499,
+      "step": 500
+    },
+    {
+      "epoch": 1.1902497027348395,
+      "grad_norm": 1.0171080592281492,
+      "learning_rate": 1.6347634071273932e-05,
+      "loss": 1.9646,
+      "step": 501
+    },
+    {
+      "epoch": 1.192627824019025,
+      "grad_norm": 1.0777884529528277,
+      "learning_rate": 1.6332820426369567e-05,
+      "loss": 2.0598,
+      "step": 502
+    },
+    {
+      "epoch": 1.1950059453032105,
+      "grad_norm": 1.020533123546286,
+      "learning_rate": 1.631798354291179e-05,
+      "loss": 1.8633,
+      "step": 503
+    },
+    {
+      "epoch": 1.197384066587396,
+      "grad_norm": 1.2067764605837605,
+      "learning_rate": 1.6303123475345182e-05,
+      "loss": 1.9119,
+      "step": 504
+    },
+    {
+      "epoch": 1.1997621878715814,
+      "grad_norm": 1.053791154347935,
+      "learning_rate": 1.6288240278199393e-05,
+      "loss": 2.1143,
+      "step": 505
+    },
+    {
+      "epoch": 1.202140309155767,
+      "grad_norm": 1.0193158748451816,
+      "learning_rate": 1.6273334006088946e-05,
+      "loss": 1.9422,
+      "step": 506
+    },
+    {
+      "epoch": 1.2045184304399523,
+      "grad_norm": 1.1993210309578368,
+      "learning_rate": 1.6258404713713035e-05,
+      "loss": 1.9505,
+      "step": 507
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 0.9278344614652998,
+      "learning_rate": 1.624345245585534e-05,
+      "loss": 1.8948,
+      "step": 508
+    },
+    {
+      "epoch": 1.2092746730083235,
+      "grad_norm": 0.8732547686659468,
+      "learning_rate": 1.62284772873838e-05,
+      "loss": 1.8943,
+      "step": 509
+    },
+    {
+      "epoch": 1.2116527942925088,
+      "grad_norm": 1.1985862989626987,
+      "learning_rate": 1.6213479263250433e-05,
+      "loss": 1.8923,
+      "step": 510
+    },
+    {
+      "epoch": 1.2140309155766944,
+      "grad_norm": 0.9559917937082356,
+      "learning_rate": 1.6198458438491124e-05,
+      "loss": 1.9478,
+      "step": 511
+    },
+    {
+      "epoch": 1.21640903686088,
+      "grad_norm": 1.0369642986640548,
+      "learning_rate": 1.6183414868225434e-05,
+      "loss": 1.932,
+      "step": 512
+    },
+    {
+      "epoch": 1.2187871581450653,
+      "grad_norm": 0.8541777304922422,
+      "learning_rate": 1.616834860765637e-05,
+      "loss": 2.0197,
+      "step": 513
+    },
+    {
+      "epoch": 1.221165279429251,
+      "grad_norm": 1.1533492531760159,
+      "learning_rate": 1.6153259712070225e-05,
+      "loss": 2.0733,
+      "step": 514
+    },
+    {
+      "epoch": 1.2235434007134365,
+      "grad_norm": 1.1544430837733473,
+      "learning_rate": 1.613814823683634e-05,
+      "loss": 1.9616,
+      "step": 515
+    },
+    {
+      "epoch": 1.2259215219976218,
+      "grad_norm": 0.904111613570178,
+      "learning_rate": 1.6123014237406912e-05,
+      "loss": 2.1678,
+      "step": 516
+    },
+    {
+      "epoch": 1.2282996432818074,
+      "grad_norm": 1.101310926122992,
+      "learning_rate": 1.6107857769316798e-05,
+      "loss": 1.8998,
+      "step": 517
+    },
+    {
+      "epoch": 1.2306777645659928,
+      "grad_norm": 0.9065269772612409,
+      "learning_rate": 1.6092678888183298e-05,
+      "loss": 2.0495,
+      "step": 518
+    },
+    {
+      "epoch": 1.2330558858501783,
+      "grad_norm": 1.252782833936401,
+      "learning_rate": 1.6077477649705963e-05,
+      "loss": 2.1285,
+      "step": 519
+    },
+    {
+      "epoch": 1.235434007134364,
+      "grad_norm": 0.9769860406730624,
+      "learning_rate": 1.6062254109666383e-05,
+      "loss": 2.0681,
+      "step": 520
+    },
+    {
+      "epoch": 1.2378121284185493,
+      "grad_norm": 1.1110445238181776,
+      "learning_rate": 1.604700832392798e-05,
+      "loss": 1.8933,
+      "step": 521
+    },
+    {
+      "epoch": 1.2401902497027348,
+      "grad_norm": 1.044536874093627,
+      "learning_rate": 1.603174034843582e-05,
+      "loss": 1.932,
+      "step": 522
+    },
+    {
+      "epoch": 1.2425683709869204,
+      "grad_norm": 1.271022221507985,
+      "learning_rate": 1.601645023921638e-05,
+      "loss": 2.0284,
+      "step": 523
+    },
+    {
+      "epoch": 1.2449464922711058,
+      "grad_norm": 1.1152833273519336,
+      "learning_rate": 1.600113805237737e-05,
+      "loss": 1.9726,
+      "step": 524
+    },
+    {
+      "epoch": 1.2473246135552913,
+      "grad_norm": 1.2380069046576747,
+      "learning_rate": 1.5985803844107502e-05,
+      "loss": 1.8106,
+      "step": 525
+    },
+    {
+      "epoch": 1.2473246135552913,
+      "eval_loss": 2.4183852672576904,
+      "eval_runtime": 65.6332,
+      "eval_samples_per_second": 9.736,
+      "eval_steps_per_second": 1.219,
+      "step": 525
+    },
+    {
+      "epoch": 1.249702734839477,
+      "grad_norm": 1.4005597471855011,
+      "learning_rate": 1.5970447670676314e-05,
+      "loss": 1.9685,
+      "step": 526
+    },
+    {
+      "epoch": 1.2520808561236623,
+      "grad_norm": 0.9054766632630273,
+      "learning_rate": 1.5955069588433932e-05,
+      "loss": 1.9919,
+      "step": 527
+    },
+    {
+      "epoch": 1.2544589774078478,
+      "grad_norm": 1.1186292190334863,
+      "learning_rate": 1.5939669653810882e-05,
+      "loss": 1.8359,
+      "step": 528
+    },
+    {
+      "epoch": 1.2568370986920332,
+      "grad_norm": 0.8661419871842052,
+      "learning_rate": 1.5924247923317874e-05,
+      "loss": 2.0698,
+      "step": 529
+    },
+    {
+      "epoch": 1.2592152199762188,
+      "grad_norm": 1.3186144918707836,
+      "learning_rate": 1.5908804453545608e-05,
+      "loss": 1.8845,
+      "step": 530
+    },
+    {
+      "epoch": 1.2615933412604043,
+      "grad_norm": 1.3508521915293898,
+      "learning_rate": 1.589333930116455e-05,
+      "loss": 2.0008,
+      "step": 531
+    },
+    {
+      "epoch": 1.2639714625445897,
+      "grad_norm": 1.0426903382069417,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 2.14,
+      "step": 532
+    },
+    {
+      "epoch": 1.2663495838287753,
+      "grad_norm": 1.1270060025806774,
+      "learning_rate": 1.5862344175655547e-05,
+      "loss": 1.7585,
+      "step": 533
+    },
+    {
+      "epoch": 1.2687277051129606,
+      "grad_norm": 0.894696959523861,
+      "learning_rate": 1.584681431626553e-05,
+      "loss": 1.7954,
+      "step": 534
+    },
+    {
+      "epoch": 1.2711058263971462,
+      "grad_norm": 0.8963443148345133,
+      "learning_rate": 1.5831263001742168e-05,
+      "loss": 1.8992,
+      "step": 535
+    },
+    {
+      "epoch": 1.2734839476813318,
+      "grad_norm": 0.9751034618814204,
+      "learning_rate": 1.581569028915166e-05,
+      "loss": 2.0839,
+      "step": 536
+    },
+    {
+      "epoch": 1.2758620689655173,
+      "grad_norm": 1.2846795383089968,
+      "learning_rate": 1.5800096235638735e-05,
+      "loss": 2.0306,
+      "step": 537
+    },
+    {
+      "epoch": 1.2782401902497027,
+      "grad_norm": 1.1344265063076209,
+      "learning_rate": 1.5784480898426444e-05,
+      "loss": 2.0588,
+      "step": 538
+    },
+    {
+      "epoch": 1.2806183115338883,
+      "grad_norm": 2.3160794394882807,
+      "learning_rate": 1.5768844334815923e-05,
+      "loss": 2.2461,
+      "step": 539
+    },
+    {
+      "epoch": 1.2829964328180736,
+      "grad_norm": 1.2198804497684061,
+      "learning_rate": 1.5753186602186207e-05,
+      "loss": 2.0165,
+      "step": 540
+    },
+    {
+      "epoch": 1.2853745541022592,
+      "grad_norm": 1.0333157223752394,
+      "learning_rate": 1.573750775799401e-05,
+      "loss": 1.9158,
+      "step": 541
+    },
+    {
+      "epoch": 1.2877526753864448,
+      "grad_norm": 1.0933945363135282,
+      "learning_rate": 1.572180785977352e-05,
+      "loss": 2.0026,
+      "step": 542
+    },
+    {
+      "epoch": 1.2901307966706301,
+      "grad_norm": 1.0383257582502903,
+      "learning_rate": 1.570608696513618e-05,
+      "loss": 2.1656,
+      "step": 543
+    },
+    {
+      "epoch": 1.2925089179548157,
+      "grad_norm": 1.238965226787101,
+      "learning_rate": 1.5690345131770474e-05,
+      "loss": 1.8335,
+      "step": 544
+    },
+    {
+      "epoch": 1.294887039239001,
+      "grad_norm": 0.7431372863827379,
+      "learning_rate": 1.5674582417441734e-05,
+      "loss": 2.0428,
+      "step": 545
+    },
+    {
+      "epoch": 1.2972651605231866,
+      "grad_norm": 1.482428714196644,
+      "learning_rate": 1.5658798879991905e-05,
+      "loss": 2.0023,
+      "step": 546
+    },
+    {
+      "epoch": 1.2996432818073722,
+      "grad_norm": 1.2061227422514962,
+      "learning_rate": 1.564299457733935e-05,
+      "loss": 2.0852,
+      "step": 547
+    },
+    {
+      "epoch": 1.3020214030915578,
+      "grad_norm": 1.2166153788730312,
+      "learning_rate": 1.5627169567478627e-05,
+      "loss": 1.7886,
+      "step": 548
+    },
+    {
+      "epoch": 1.3043995243757431,
+      "grad_norm": 0.9980529341355449,
+      "learning_rate": 1.561132390848029e-05,
+      "loss": 1.8238,
+      "step": 549
+    },
+    {
+      "epoch": 1.3067776456599287,
+      "grad_norm": 1.0516168634342369,
+      "learning_rate": 1.5595457658490643e-05,
+      "loss": 1.9991,
+      "step": 550
+    },
+    {
+      "epoch": 1.309155766944114,
+      "grad_norm": 1.1353824503332424,
+      "learning_rate": 1.5579570875731572e-05,
+      "loss": 2.2081,
+      "step": 551
+    },
+    {
+      "epoch": 1.3115338882282996,
+      "grad_norm": 1.0157342209528704,
+      "learning_rate": 1.55636636185003e-05,
+      "loss": 1.9142,
+      "step": 552
+    },
+    {
+      "epoch": 1.3139120095124852,
+      "grad_norm": 1.3048380581725043,
+      "learning_rate": 1.5547735945169188e-05,
+      "loss": 1.9015,
+      "step": 553
+    },
+    {
+      "epoch": 1.3162901307966706,
+      "grad_norm": 1.0849291353506971,
+      "learning_rate": 1.55317879141855e-05,
+      "loss": 1.892,
+      "step": 554
+    },
+    {
+      "epoch": 1.3186682520808561,
+      "grad_norm": 1.103994377178752,
+      "learning_rate": 1.5515819584071216e-05,
+      "loss": 2.0736,
+      "step": 555
+    },
+    {
+      "epoch": 1.3210463733650415,
+      "grad_norm": 0.87333743581362,
+      "learning_rate": 1.5499831013422804e-05,
+      "loss": 1.6585,
+      "step": 556
+    },
+    {
+      "epoch": 1.323424494649227,
+      "grad_norm": 1.5096036417812366,
+      "learning_rate": 1.5483822260911002e-05,
+      "loss": 2.0984,
+      "step": 557
+    },
+    {
+      "epoch": 1.3258026159334126,
+      "grad_norm": 0.9402021272644058,
+      "learning_rate": 1.5467793385280602e-05,
+      "loss": 1.919,
+      "step": 558
+    },
+    {
+      "epoch": 1.3281807372175982,
+      "grad_norm": 1.010147060155673,
+      "learning_rate": 1.5451744445350252e-05,
+      "loss": 1.789,
+      "step": 559
+    },
+    {
+      "epoch": 1.3305588585017836,
+      "grad_norm": 1.018932157068251,
+      "learning_rate": 1.5435675500012212e-05,
+      "loss": 1.9567,
+      "step": 560
+    },
+    {
+      "epoch": 1.3329369797859691,
+      "grad_norm": 0.9622353435965788,
+      "learning_rate": 1.5419586608232163e-05,
+      "loss": 2.102,
+      "step": 561
+    },
+    {
+      "epoch": 1.3353151010701545,
+      "grad_norm": 1.046337850818545,
+      "learning_rate": 1.540347782904897e-05,
+      "loss": 1.8589,
+      "step": 562
+    },
+    {
+      "epoch": 1.33769322235434,
+      "grad_norm": 1.0106090373600465,
+      "learning_rate": 1.5387349221574493e-05,
+      "loss": 2.0144,
+      "step": 563
+    },
+    {
+      "epoch": 1.3400713436385256,
+      "grad_norm": 1.1400911418818147,
+      "learning_rate": 1.5371200844993332e-05,
+      "loss": 1.9842,
+      "step": 564
+    },
+    {
+      "epoch": 1.342449464922711,
+      "grad_norm": 0.8684778766824857,
+      "learning_rate": 1.535503275856264e-05,
+      "loss": 2.0641,
+      "step": 565
+    },
+    {
+      "epoch": 1.3448275862068966,
+      "grad_norm": 0.9355737090349231,
+      "learning_rate": 1.5338845021611906e-05,
+      "loss": 1.8084,
+      "step": 566
+    },
+    {
+      "epoch": 1.347205707491082,
+      "grad_norm": 1.0306221444200083,
+      "learning_rate": 1.5322637693542716e-05,
+      "loss": 2.0131,
+      "step": 567
+    },
+    {
+      "epoch": 1.3495838287752675,
+      "grad_norm": 1.0214723689545668,
+      "learning_rate": 1.5306410833828534e-05,
+      "loss": 1.969,
+      "step": 568
+    },
+    {
+      "epoch": 1.351961950059453,
+      "grad_norm": 0.9625636615112916,
+      "learning_rate": 1.529016450201453e-05,
+      "loss": 1.9543,
+      "step": 569
+    },
+    {
+      "epoch": 1.3543400713436387,
+      "grad_norm": 0.9724729717579162,
+      "learning_rate": 1.5273898757717295e-05,
+      "loss": 1.929,
+      "step": 570
+    },
+    {
+      "epoch": 1.356718192627824,
+      "grad_norm": 1.0109447015124593,
+      "learning_rate": 1.5257613660624673e-05,
+      "loss": 1.9279,
+      "step": 571
+    },
+    {
+      "epoch": 1.3590963139120096,
+      "grad_norm": 1.008542064536025,
+      "learning_rate": 1.5241309270495524e-05,
+      "loss": 2.0551,
+      "step": 572
+    },
+    {
+      "epoch": 1.361474435196195,
+      "grad_norm": 1.2854483484165498,
+      "learning_rate": 1.5224985647159489e-05,
+      "loss": 2.0914,
+      "step": 573
+    },
+    {
+      "epoch": 1.3638525564803805,
+      "grad_norm": 0.9055884973780003,
+      "learning_rate": 1.5208642850516806e-05,
+      "loss": 1.8825,
+      "step": 574
+    },
+    {
+      "epoch": 1.366230677764566,
+      "grad_norm": 1.0375062326116673,
+      "learning_rate": 1.5192280940538058e-05,
+      "loss": 1.8674,
+      "step": 575
+    },
+    {
+      "epoch": 1.3686087990487514,
+      "grad_norm": 0.896010533212188,
+      "learning_rate": 1.5175899977263963e-05,
+      "loss": 2.0912,
+      "step": 576
+    },
+    {
+      "epoch": 1.370986920332937,
+      "grad_norm": 1.1501881622163808,
+      "learning_rate": 1.5159500020805173e-05,
+      "loss": 1.89,
+      "step": 577
+    },
+    {
+      "epoch": 1.3733650416171224,
+      "grad_norm": 0.9783540504581939,
+      "learning_rate": 1.5143081131342015e-05,
+      "loss": 1.9331,
+      "step": 578
+    },
+    {
+      "epoch": 1.375743162901308,
+      "grad_norm": 1.170097731492715,
+      "learning_rate": 1.5126643369124303e-05,
+      "loss": 2.0069,
+      "step": 579
+    },
+    {
+      "epoch": 1.3781212841854935,
+      "grad_norm": 1.0072974136421486,
+      "learning_rate": 1.5110186794471105e-05,
+      "loss": 2.0293,
+      "step": 580
+    },
+    {
+      "epoch": 1.380499405469679,
+      "grad_norm": 1.1697150901196478,
+      "learning_rate": 1.5093711467770517e-05,
+      "loss": 1.9685,
+      "step": 581
+    },
+    {
+      "epoch": 1.3828775267538644,
+      "grad_norm": 1.2206930372756868,
+      "learning_rate": 1.5077217449479455e-05,
+      "loss": 1.933,
+      "step": 582
+    },
+    {
+      "epoch": 1.38525564803805,
+      "grad_norm": 0.970392976217639,
+      "learning_rate": 1.5060704800123413e-05,
+      "loss": 2.0112,
+      "step": 583
+    },
+    {
+      "epoch": 1.3876337693222354,
+      "grad_norm": 1.0661843012528456,
+      "learning_rate": 1.5044173580296267e-05,
+      "loss": 1.9158,
+      "step": 584
+    },
+    {
+      "epoch": 1.390011890606421,
+      "grad_norm": 0.8514185489462422,
+      "learning_rate": 1.502762385066002e-05,
+      "loss": 1.8758,
+      "step": 585
+    },
+    {
+      "epoch": 1.3923900118906065,
+      "grad_norm": 0.9748541634753175,
+      "learning_rate": 1.5011055671944616e-05,
+      "loss": 2.1468,
+      "step": 586
+    },
+    {
+      "epoch": 1.3947681331747919,
+      "grad_norm": 0.8818979397080599,
+      "learning_rate": 1.4994469104947686e-05,
+      "loss": 1.9446,
+      "step": 587
+    },
+    {
+      "epoch": 1.3971462544589774,
+      "grad_norm": 1.167689162690369,
+      "learning_rate": 1.4977864210534341e-05,
+      "loss": 1.706,
+      "step": 588
+    },
+    {
+      "epoch": 1.3995243757431628,
+      "grad_norm": 0.9558567418999584,
+      "learning_rate": 1.496124104963695e-05,
+      "loss": 2.1427,
+      "step": 589
+    },
+    {
+      "epoch": 1.4019024970273484,
+      "grad_norm": 0.991232122819226,
+      "learning_rate": 1.4944599683254903e-05,
+      "loss": 2.0682,
+      "step": 590
+    },
+    {
+      "epoch": 1.404280618311534,
+      "grad_norm": 1.0316875876904197,
+      "learning_rate": 1.4927940172454405e-05,
+      "loss": 2.1085,
+      "step": 591
+    },
+    {
+      "epoch": 1.4066587395957193,
+      "grad_norm": 1.0732162496383806,
+      "learning_rate": 1.4911262578368233e-05,
+      "loss": 1.8279,
+      "step": 592
+    },
+    {
+      "epoch": 1.4090368608799049,
+      "grad_norm": 0.9257277902304547,
+      "learning_rate": 1.4894566962195532e-05,
+      "loss": 2.0153,
+      "step": 593
+    },
+    {
+      "epoch": 1.4114149821640904,
+      "grad_norm": 1.0669817139188487,
+      "learning_rate": 1.4877853385201569e-05,
+      "loss": 1.9106,
+      "step": 594
+    },
+    {
+      "epoch": 1.4137931034482758,
+      "grad_norm": 0.8934624324805909,
+      "learning_rate": 1.4861121908717529e-05,
+      "loss": 1.9279,
+      "step": 595
+    },
+    {
+      "epoch": 1.4161712247324614,
+      "grad_norm": 1.0784062752165995,
+      "learning_rate": 1.4844372594140271e-05,
+      "loss": 1.8346,
+      "step": 596
+    },
+    {
+      "epoch": 1.418549346016647,
+      "grad_norm": 1.0964764861814489,
+      "learning_rate": 1.4827605502932118e-05,
+      "loss": 2.1296,
+      "step": 597
+    },
+    {
+      "epoch": 1.4209274673008323,
+      "grad_norm": 15.356619023384987,
+      "learning_rate": 1.4810820696620625e-05,
+      "loss": 2.1312,
+      "step": 598
+    },
+    {
+      "epoch": 1.4233055885850179,
+      "grad_norm": 1.1118633779533458,
+      "learning_rate": 1.4794018236798349e-05,
+      "loss": 1.8619,
+      "step": 599
+    },
+    {
+      "epoch": 1.4256837098692032,
+      "grad_norm": 1.153501655154434,
+      "learning_rate": 1.477719818512263e-05,
+      "loss": 2.077,
+      "step": 600
+    },
+    {
+      "epoch": 1.4280618311533888,
+      "grad_norm": 1.0076021147835075,
+      "learning_rate": 1.4760360603315362e-05,
+      "loss": 1.9109,
+      "step": 601
+    },
+    {
+      "epoch": 1.4304399524375744,
+      "grad_norm": 0.9674703722137163,
+      "learning_rate": 1.4743505553162765e-05,
+      "loss": 2.0096,
+      "step": 602
+    },
+    {
+      "epoch": 1.4328180737217597,
+      "grad_norm": 1.0412828169607384,
+      "learning_rate": 1.4726633096515163e-05,
+      "loss": 1.9162,
+      "step": 603
+    },
+    {
+      "epoch": 1.4351961950059453,
+      "grad_norm": 0.8719798048430073,
+      "learning_rate": 1.4709743295286751e-05,
+      "loss": 1.9538,
+      "step": 604
+    },
+    {
+      "epoch": 1.4375743162901307,
+      "grad_norm": 1.1442918189307298,
+      "learning_rate": 1.4692836211455373e-05,
+      "loss": 1.9587,
+      "step": 605
+    },
+    {
+      "epoch": 1.4399524375743162,
+      "grad_norm": 1.023556986586939,
+      "learning_rate": 1.4675911907062289e-05,
+      "loss": 1.8141,
+      "step": 606
+    },
+    {
+      "epoch": 1.4423305588585018,
+      "grad_norm": 0.8486875820991789,
+      "learning_rate": 1.4658970444211953e-05,
+      "loss": 1.9474,
+      "step": 607
+    },
+    {
+      "epoch": 1.4447086801426874,
+      "grad_norm": 0.906460047909353,
+      "learning_rate": 1.464201188507178e-05,
+      "loss": 2.2903,
+      "step": 608
+    },
+    {
+      "epoch": 1.4470868014268727,
+      "grad_norm": 1.1791852257245115,
+      "learning_rate": 1.4625036291871926e-05,
+      "loss": 2.0229,
+      "step": 609
+    },
+    {
+      "epoch": 1.4494649227110583,
+      "grad_norm": 1.051198365641329,
+      "learning_rate": 1.460804372690505e-05,
+      "loss": 1.8857,
+      "step": 610
+    },
+    {
+      "epoch": 1.4518430439952437,
+      "grad_norm": 0.8672989481026866,
+      "learning_rate": 1.4591034252526093e-05,
+      "loss": 1.9414,
+      "step": 611
+    },
+    {
+      "epoch": 1.4542211652794292,
+      "grad_norm": 0.9235870505947725,
+      "learning_rate": 1.4574007931152037e-05,
+      "loss": 2.0249,
+      "step": 612
+    },
+    {
+      "epoch": 1.4565992865636148,
+      "grad_norm": 1.0522141111911232,
+      "learning_rate": 1.4556964825261696e-05,
+      "loss": 2.2079,
+      "step": 613
+    },
+    {
+      "epoch": 1.4589774078478002,
+      "grad_norm": 1.006897151013655,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 1.9247,
+      "step": 614
+    },
+    {
+      "epoch": 1.4613555291319857,
+      "grad_norm": 0.9494032663031484,
+      "learning_rate": 1.4522828510155123e-05,
+      "loss": 1.9099,
+      "step": 615
+    },
+    {
+      "epoch": 1.463733650416171,
+      "grad_norm": 1.0076075073121573,
+      "learning_rate": 1.4505735426203545e-05,
+      "loss": 1.8666,
+      "step": 616
+    },
+    {
+      "epoch": 1.4661117717003567,
+      "grad_norm": 1.1408823580186684,
+      "learning_rate": 1.4488625808264536e-05,
+      "loss": 1.9044,
+      "step": 617
+    },
+    {
+      "epoch": 1.4684898929845422,
+      "grad_norm": 0.9085119845888443,
+      "learning_rate": 1.4471499719122565e-05,
+      "loss": 1.8932,
+      "step": 618
+    },
+    {
+      "epoch": 1.4708680142687278,
+      "grad_norm": 1.148726123633719,
+      "learning_rate": 1.4454357221622546e-05,
+      "loss": 1.9092,
+      "step": 619
+    },
+    {
+      "epoch": 1.4732461355529132,
+      "grad_norm": 0.9671542619932931,
+      "learning_rate": 1.4437198378669598e-05,
+      "loss": 1.9493,
+      "step": 620
+    },
+    {
+      "epoch": 1.4756242568370987,
+      "grad_norm": 1.2173093104656718,
+      "learning_rate": 1.4420023253228825e-05,
+      "loss": 2.2601,
+      "step": 621
+    },
+    {
+      "epoch": 1.478002378121284,
+      "grad_norm": 1.1927753831747965,
+      "learning_rate": 1.4402831908325082e-05,
+      "loss": 1.8732,
+      "step": 622
+    },
+    {
+      "epoch": 1.4803804994054697,
+      "grad_norm": 1.0640642993825382,
+      "learning_rate": 1.4385624407042729e-05,
+      "loss": 2.0229,
+      "step": 623
+    },
+    {
+      "epoch": 1.4827586206896552,
+      "grad_norm": 1.0742996295504268,
+      "learning_rate": 1.4368400812525434e-05,
+      "loss": 1.8983,
+      "step": 624
+    },
+    {
+      "epoch": 1.4851367419738406,
+      "grad_norm": 0.9228290280482974,
+      "learning_rate": 1.4351161187975904e-05,
+      "loss": 2.1249,
+      "step": 625
+    },
+    {
+      "epoch": 1.4875148632580262,
+      "grad_norm": 0.9738223203214996,
+      "learning_rate": 1.4333905596655668e-05,
+      "loss": 1.9615,
+      "step": 626
+    },
+    {
+      "epoch": 1.4898929845422115,
+      "grad_norm": 1.0847722352198572,
+      "learning_rate": 1.4316634101884859e-05,
+      "loss": 2.0659,
+      "step": 627
+    },
+    {
+      "epoch": 1.492271105826397,
+      "grad_norm": 1.2154947037501092,
+      "learning_rate": 1.4299346767041956e-05,
+      "loss": 2.0163,
+      "step": 628
+    },
+    {
+      "epoch": 1.4946492271105827,
+      "grad_norm": 1.1111031435058485,
+      "learning_rate": 1.4282043655563566e-05,
+      "loss": 2.1607,
+      "step": 629
+    },
+    {
+      "epoch": 1.4970273483947683,
+      "grad_norm": 1.225646636267743,
+      "learning_rate": 1.4264724830944198e-05,
+      "loss": 2.0063,
+      "step": 630
+    },
+    {
+      "epoch": 1.4970273483947683,
+      "eval_loss": 2.4099409580230713,
+      "eval_runtime": 65.5151,
+      "eval_samples_per_second": 9.753,
+      "eval_steps_per_second": 1.221,
+      "step": 630
+    },
+    {
+      "epoch": 1.4994054696789536,
+      "grad_norm": 0.9404929651596697,
+      "learning_rate": 1.4247390356736011e-05,
+      "loss": 2.1291,
+      "step": 631
+    },
+    {
+      "epoch": 1.501783590963139,
+      "grad_norm": 1.166127163010488,
+      "learning_rate": 1.4230040296548588e-05,
+      "loss": 1.9927,
+      "step": 632
+    },
+    {
+      "epoch": 1.5041617122473245,
+      "grad_norm": 0.9098808290926231,
+      "learning_rate": 1.4212674714048721e-05,
+      "loss": 2.0014,
+      "step": 633
+    },
+    {
+      "epoch": 1.50653983353151,
+      "grad_norm": 1.2163691555238052,
+      "learning_rate": 1.4195293672960148e-05,
+      "loss": 1.8449,
+      "step": 634
+    },
+    {
+      "epoch": 1.5089179548156957,
+      "grad_norm": 1.0668500527745948,
+      "learning_rate": 1.4177897237063336e-05,
+      "loss": 1.8293,
+      "step": 635
+    },
+    {
+      "epoch": 1.5112960760998813,
+      "grad_norm": 0.9965754757162669,
+      "learning_rate": 1.4160485470195245e-05,
+      "loss": 1.9292,
+      "step": 636
+    },
+    {
+      "epoch": 1.5136741973840666,
+      "grad_norm": 0.9613885332924135,
+      "learning_rate": 1.4143058436249095e-05,
+      "loss": 1.9555,
+      "step": 637
+    },
+    {
+      "epoch": 1.516052318668252,
+      "grad_norm": 1.029262128237135,
+      "learning_rate": 1.4125616199174125e-05,
+      "loss": 2.0374,
+      "step": 638
+    },
+    {
+      "epoch": 1.5184304399524375,
+      "grad_norm": 1.2261767783176603,
+      "learning_rate": 1.4108158822975368e-05,
+      "loss": 1.8905,
+      "step": 639
+    },
+    {
+      "epoch": 1.5208085612366231,
+      "grad_norm": 1.1691866000837525,
+      "learning_rate": 1.4090686371713403e-05,
+      "loss": 2.0864,
+      "step": 640
+    },
+    {
+      "epoch": 1.5231866825208087,
+      "grad_norm": 1.269895781196831,
+      "learning_rate": 1.4073198909504128e-05,
+      "loss": 1.9734,
+      "step": 641
+    },
+    {
+      "epoch": 1.525564803804994,
+      "grad_norm": 1.2954844306148612,
+      "learning_rate": 1.4055696500518539e-05,
+      "loss": 1.8801,
+      "step": 642
+    },
+    {
+      "epoch": 1.5279429250891794,
+      "grad_norm": 1.056075124876724,
+      "learning_rate": 1.4038179208982459e-05,
+      "loss": 1.9915,
+      "step": 643
+    },
+    {
+      "epoch": 1.530321046373365,
+      "grad_norm": 1.2282288043296659,
+      "learning_rate": 1.402064709917634e-05,
+      "loss": 1.9476,
+      "step": 644
+    },
+    {
+      "epoch": 1.5326991676575505,
+      "grad_norm": 1.3379065950112186,
+      "learning_rate": 1.4003100235434998e-05,
+      "loss": 2.0888,
+      "step": 645
+    },
+    {
+      "epoch": 1.5350772889417361,
+      "grad_norm": 0.8299776495667625,
+      "learning_rate": 1.3985538682147395e-05,
+      "loss": 1.9002,
+      "step": 646
+    },
+    {
+      "epoch": 1.5374554102259215,
+      "grad_norm": 1.3323866450270925,
+      "learning_rate": 1.3967962503756401e-05,
+      "loss": 1.8784,
+      "step": 647
+    },
+    {
+      "epoch": 1.539833531510107,
+      "grad_norm": 1.2995832056920724,
+      "learning_rate": 1.3950371764758543e-05,
+      "loss": 1.687,
+      "step": 648
+    },
+    {
+      "epoch": 1.5422116527942924,
+      "grad_norm": 1.0444385989486211,
+      "learning_rate": 1.3932766529703787e-05,
+      "loss": 2.0069,
+      "step": 649
+    },
+    {
+      "epoch": 1.544589774078478,
+      "grad_norm": 1.0757353584625595,
+      "learning_rate": 1.3915146863195292e-05,
+      "loss": 1.9444,
+      "step": 650
+    },
+    {
+      "epoch": 1.5469678953626635,
+      "grad_norm": 1.011252037188437,
+      "learning_rate": 1.3897512829889168e-05,
+      "loss": 1.8028,
+      "step": 651
+    },
+    {
+      "epoch": 1.5493460166468491,
+      "grad_norm": 1.0760288809230223,
+      "learning_rate": 1.3879864494494252e-05,
+      "loss": 2.0473,
+      "step": 652
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 1.242394869159288,
+      "learning_rate": 1.3862201921771864e-05,
+      "loss": 2.1959,
+      "step": 653
+    },
+    {
+      "epoch": 1.5541022592152198,
+      "grad_norm": 1.052411681271537,
+      "learning_rate": 1.3844525176535557e-05,
+      "loss": 2.0419,
+      "step": 654
+    },
+    {
+      "epoch": 1.5564803804994054,
+      "grad_norm": 0.9881132820543386,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 1.8639,
+      "step": 655
+    },
+    {
+      "epoch": 1.558858501783591,
+      "grad_norm": 1.1464153070903584,
+      "learning_rate": 1.3809129428035229e-05,
+      "loss": 1.7864,
+      "step": 656
+    },
+    {
+      "epoch": 1.5612366230677766,
+      "grad_norm": 0.9309263598280921,
+      "learning_rate": 1.3791410554657412e-05,
+      "loss": 1.9641,
+      "step": 657
+    },
+    {
+      "epoch": 1.563614744351962,
+      "grad_norm": 1.0278232179656681,
+      "learning_rate": 1.3773677768537608e-05,
+      "loss": 2.0758,
+      "step": 658
+    },
+    {
+      "epoch": 1.5659928656361475,
+      "grad_norm": 1.076577593564504,
+      "learning_rate": 1.3755931134747024e-05,
+      "loss": 2.0546,
+      "step": 659
+    },
+    {
+      "epoch": 1.5683709869203328,
+      "grad_norm": 1.10395359678112,
+      "learning_rate": 1.3738170718407689e-05,
+      "loss": 1.9597,
+      "step": 660
+    },
+    {
+      "epoch": 1.5707491082045184,
+      "grad_norm": 1.0709571819780048,
+      "learning_rate": 1.3720396584692204e-05,
+      "loss": 1.8091,
+      "step": 661
+    },
+    {
+      "epoch": 1.573127229488704,
+      "grad_norm": 0.9361930895015704,
+      "learning_rate": 1.3702608798823506e-05,
+      "loss": 2.151,
+      "step": 662
+    },
+    {
+      "epoch": 1.5755053507728896,
+      "grad_norm": 0.960233741717821,
+      "learning_rate": 1.3684807426074637e-05,
+      "loss": 2.0959,
+      "step": 663
+    },
+    {
+      "epoch": 1.577883472057075,
+      "grad_norm": 1.0144097356398964,
+      "learning_rate": 1.3666992531768482e-05,
+      "loss": 2.0612,
+      "step": 664
+    },
+    {
+      "epoch": 1.5802615933412603,
+      "grad_norm": 1.0093128761118164,
+      "learning_rate": 1.3649164181277554e-05,
+      "loss": 1.9669,
+      "step": 665
+    },
+    {
+      "epoch": 1.5826397146254458,
+      "grad_norm": 0.8991548607233194,
+      "learning_rate": 1.3631322440023743e-05,
+      "loss": 1.9549,
+      "step": 666
+    },
+    {
+      "epoch": 1.5850178359096314,
+      "grad_norm": 1.1136062306803636,
+      "learning_rate": 1.361346737347808e-05,
+      "loss": 1.7156,
+      "step": 667
+    },
+    {
+      "epoch": 1.587395957193817,
+      "grad_norm": 1.0226580218095702,
+      "learning_rate": 1.359559904716048e-05,
+      "loss": 2.0015,
+      "step": 668
+    },
+    {
+      "epoch": 1.5897740784780023,
+      "grad_norm": 1.0629689645649965,
+      "learning_rate": 1.3577717526639532e-05,
+      "loss": 1.9631,
+      "step": 669
+    },
+    {
+      "epoch": 1.592152199762188,
+      "grad_norm": 0.9835886516605551,
+      "learning_rate": 1.3559822877532234e-05,
+      "loss": 2.0375,
+      "step": 670
+    },
+    {
+      "epoch": 1.5945303210463733,
+      "grad_norm": 1.0706093353458817,
+      "learning_rate": 1.354191516550376e-05,
+      "loss": 2.1751,
+      "step": 671
+    },
+    {
+      "epoch": 1.5969084423305588,
+      "grad_norm": 0.9758001603671903,
+      "learning_rate": 1.352399445626722e-05,
+      "loss": 2.0163,
+      "step": 672
+    },
+    {
+      "epoch": 1.5992865636147444,
+      "grad_norm": 0.9532313551731908,
+      "learning_rate": 1.3506060815583415e-05,
+      "loss": 1.9762,
+      "step": 673
+    },
+    {
+      "epoch": 1.60166468489893,
+      "grad_norm": 1.1096343614759168,
+      "learning_rate": 1.3488114309260603e-05,
+      "loss": 1.9759,
+      "step": 674
+    },
+    {
+      "epoch": 1.6040428061831153,
+      "grad_norm": 0.9157371697281291,
+      "learning_rate": 1.3470155003154251e-05,
+      "loss": 2.0912,
+      "step": 675
+    },
+    {
+      "epoch": 1.6064209274673007,
+      "grad_norm": 1.0482149472996887,
+      "learning_rate": 1.3452182963166792e-05,
+      "loss": 2.1044,
+      "step": 676
+    },
+    {
+      "epoch": 1.6087990487514863,
+      "grad_norm": 0.8905172826303334,
+      "learning_rate": 1.3434198255247399e-05,
+      "loss": 1.6975,
+      "step": 677
+    },
+    {
+      "epoch": 1.6111771700356718,
+      "grad_norm": 1.058790898557101,
+      "learning_rate": 1.3416200945391711e-05,
+      "loss": 1.8314,
+      "step": 678
+    },
+    {
+      "epoch": 1.6135552913198574,
+      "grad_norm": 1.0907756047211878,
+      "learning_rate": 1.3398191099641622e-05,
+      "loss": 1.7873,
+      "step": 679
+    },
+    {
+      "epoch": 1.6159334126040428,
+      "grad_norm": 0.9443953519393393,
+      "learning_rate": 1.3380168784085028e-05,
+      "loss": 2.0068,
+      "step": 680
+    },
+    {
+      "epoch": 1.6183115338882283,
+      "grad_norm": 0.8651637816722115,
+      "learning_rate": 1.3362134064855583e-05,
+      "loss": 1.8158,
+      "step": 681
+    },
+    {
+      "epoch": 1.6206896551724137,
+      "grad_norm": 1.0280612331110308,
+      "learning_rate": 1.334408700813245e-05,
+      "loss": 2.0307,
+      "step": 682
+    },
+    {
+      "epoch": 1.6230677764565993,
+      "grad_norm": 1.1338005074791748,
+      "learning_rate": 1.3326027680140075e-05,
+      "loss": 1.884,
+      "step": 683
+    },
+    {
+      "epoch": 1.6254458977407849,
+      "grad_norm": 0.8710520136707882,
+      "learning_rate": 1.3307956147147924e-05,
+      "loss": 1.9489,
+      "step": 684
+    },
+    {
+      "epoch": 1.6278240190249704,
+      "grad_norm": 1.2459709942333668,
+      "learning_rate": 1.3289872475470257e-05,
+      "loss": 2.1625,
+      "step": 685
+    },
+    {
+      "epoch": 1.6302021403091558,
+      "grad_norm": 0.8883559924456562,
+      "learning_rate": 1.3271776731465878e-05,
+      "loss": 1.8351,
+      "step": 686
+    },
+    {
+      "epoch": 1.6325802615933411,
+      "grad_norm": 0.9532577115031986,
+      "learning_rate": 1.325366898153788e-05,
+      "loss": 1.8954,
+      "step": 687
+    },
+    {
+      "epoch": 1.6349583828775267,
+      "grad_norm": 1.0867243605338575,
+      "learning_rate": 1.3235549292133425e-05,
+      "loss": 2.023,
+      "step": 688
+    },
+    {
+      "epoch": 1.6373365041617123,
+      "grad_norm": 0.9945570973943012,
+      "learning_rate": 1.3217417729743483e-05,
+      "loss": 1.7962,
+      "step": 689
+    },
+    {
+      "epoch": 1.6397146254458979,
+      "grad_norm": 0.9349059265184032,
+      "learning_rate": 1.319927436090259e-05,
+      "loss": 2.0278,
+      "step": 690
+    },
+    {
+      "epoch": 1.6420927467300832,
+      "grad_norm": 0.9423473629997007,
+      "learning_rate": 1.3181119252188612e-05,
+      "loss": 1.9311,
+      "step": 691
+    },
+    {
+      "epoch": 1.6444708680142688,
+      "grad_norm": 1.0039996608993265,
+      "learning_rate": 1.3162952470222488e-05,
+      "loss": 1.9779,
+      "step": 692
+    },
+    {
+      "epoch": 1.6468489892984541,
+      "grad_norm": 1.0292101999229477,
+      "learning_rate": 1.3144774081667993e-05,
+      "loss": 2.0786,
+      "step": 693
+    },
+    {
+      "epoch": 1.6492271105826397,
+      "grad_norm": 0.8987803262203654,
+      "learning_rate": 1.31265841532315e-05,
+      "loss": 1.7987,
+      "step": 694
+    },
+    {
+      "epoch": 1.6516052318668253,
+      "grad_norm": 1.1875414858923983,
+      "learning_rate": 1.3108382751661722e-05,
+      "loss": 2.0059,
+      "step": 695
+    },
+    {
+      "epoch": 1.6539833531510109,
+      "grad_norm": 1.0269203634847877,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 2.0657,
+      "step": 696
+    },
+    {
+      "epoch": 1.6563614744351962,
+      "grad_norm": 1.0926982865846275,
+      "learning_rate": 1.3071945796327431e-05,
+      "loss": 2.0779,
+      "step": 697
+    },
+    {
+      "epoch": 1.6587395957193816,
+      "grad_norm": 1.0775345370788383,
+      "learning_rate": 1.3053710376269873e-05,
+      "loss": 1.9898,
+      "step": 698
+    },
+    {
+      "epoch": 1.6611177170035671,
+      "grad_norm": 1.1340932134975976,
+      "learning_rate": 1.3035463750492448e-05,
+      "loss": 2.146,
+      "step": 699
+    },
+    {
+      "epoch": 1.6634958382877527,
+      "grad_norm": 1.0731177791654507,
+      "learning_rate": 1.3017205985951926e-05,
+      "loss": 2.0155,
+      "step": 700
+    },
+    {
+      "epoch": 1.6658739595719383,
+      "grad_norm": 0.8969768193578244,
+      "learning_rate": 1.2998937149645944e-05,
+      "loss": 1.7347,
+      "step": 701
+    },
+    {
+      "epoch": 1.6682520808561236,
+      "grad_norm": 0.9835389016614261,
+      "learning_rate": 1.2980657308612778e-05,
+      "loss": 1.8539,
+      "step": 702
+    },
+    {
+      "epoch": 1.670630202140309,
+      "grad_norm": 1.1380423436787481,
+      "learning_rate": 1.2962366529931076e-05,
+      "loss": 1.9974,
+      "step": 703
+    },
+    {
+      "epoch": 1.6730083234244946,
+      "grad_norm": 1.0964744709498448,
+      "learning_rate": 1.2944064880719634e-05,
+      "loss": 1.72,
+      "step": 704
+    },
+    {
+      "epoch": 1.6753864447086801,
+      "grad_norm": 1.2519840244518319,
+      "learning_rate": 1.2925752428137126e-05,
+      "loss": 1.8856,
+      "step": 705
+    },
+    {
+      "epoch": 1.6777645659928657,
+      "grad_norm": 1.0121056685710188,
+      "learning_rate": 1.2907429239381872e-05,
+      "loss": 2.0765,
+      "step": 706
+    },
+    {
+      "epoch": 1.6801426872770513,
+      "grad_norm": 1.0586498134607165,
+      "learning_rate": 1.2889095381691594e-05,
+      "loss": 2.1218,
+      "step": 707
+    },
+    {
+      "epoch": 1.6825208085612366,
+      "grad_norm": 0.969809896629875,
+      "learning_rate": 1.287075092234316e-05,
+      "loss": 2.0508,
+      "step": 708
+    },
+    {
+      "epoch": 1.684898929845422,
+      "grad_norm": 1.0575764929431117,
+      "learning_rate": 1.2852395928652343e-05,
+      "loss": 1.9796,
+      "step": 709
+    },
+    {
+      "epoch": 1.6872770511296076,
+      "grad_norm": 0.9921956479219585,
+      "learning_rate": 1.2834030467973572e-05,
+      "loss": 2.0071,
+      "step": 710
+    },
+    {
+      "epoch": 1.6896551724137931,
+      "grad_norm": 1.0183541379675678,
+      "learning_rate": 1.2815654607699687e-05,
+      "loss": 1.9518,
+      "step": 711
+    },
+    {
+      "epoch": 1.6920332936979787,
+      "grad_norm": 1.097875724988254,
+      "learning_rate": 1.2797268415261681e-05,
+      "loss": 2.0982,
+      "step": 712
+    },
+    {
+      "epoch": 1.694411414982164,
+      "grad_norm": 0.9022331837037142,
+      "learning_rate": 1.2778871958128472e-05,
+      "loss": 1.7788,
+      "step": 713
+    },
+    {
+      "epoch": 1.6967895362663494,
+      "grad_norm": 0.8133273901988929,
+      "learning_rate": 1.2760465303806639e-05,
+      "loss": 1.8453,
+      "step": 714
+    },
+    {
+      "epoch": 1.699167657550535,
+      "grad_norm": 1.0492660319739247,
+      "learning_rate": 1.2742048519840182e-05,
+      "loss": 1.8965,
+      "step": 715
+    },
+    {
+      "epoch": 1.7015457788347206,
+      "grad_norm": 1.1192820098120624,
+      "learning_rate": 1.2723621673810277e-05,
+      "loss": 2.0887,
+      "step": 716
+    },
+    {
+      "epoch": 1.7039239001189062,
+      "grad_norm": 0.9360576963110526,
+      "learning_rate": 1.2705184833335008e-05,
+      "loss": 1.7144,
+      "step": 717
+    },
+    {
+      "epoch": 1.7063020214030915,
+      "grad_norm": 1.1130603696459762,
+      "learning_rate": 1.2686738066069148e-05,
+      "loss": 1.7416,
+      "step": 718
+    },
+    {
+      "epoch": 1.708680142687277,
+      "grad_norm": 1.0304078304302904,
+      "learning_rate": 1.2668281439703893e-05,
+      "loss": 2.1117,
+      "step": 719
+    },
+    {
+      "epoch": 1.7110582639714624,
+      "grad_norm": 1.0656197340261881,
+      "learning_rate": 1.264981502196662e-05,
+      "loss": 1.9695,
+      "step": 720
+    },
+    {
+      "epoch": 1.713436385255648,
+      "grad_norm": 1.0087951542362363,
+      "learning_rate": 1.2631338880620627e-05,
+      "loss": 1.9352,
+      "step": 721
+    },
+    {
+      "epoch": 1.7158145065398336,
+      "grad_norm": 1.0373066800416066,
+      "learning_rate": 1.2612853083464903e-05,
+      "loss": 2.0947,
+      "step": 722
+    },
+    {
+      "epoch": 1.7181926278240192,
+      "grad_norm": 1.1765591294813367,
+      "learning_rate": 1.2594357698333864e-05,
+      "loss": 2.0922,
+      "step": 723
+    },
+    {
+      "epoch": 1.7205707491082045,
+      "grad_norm": 1.1664325067262717,
+      "learning_rate": 1.2575852793097113e-05,
+      "loss": 2.1225,
+      "step": 724
+    },
+    {
+      "epoch": 1.7229488703923899,
+      "grad_norm": 1.229465533643844,
+      "learning_rate": 1.255733843565918e-05,
+      "loss": 1.861,
+      "step": 725
+    },
+    {
+      "epoch": 1.7253269916765754,
+      "grad_norm": 0.9938570044256105,
+      "learning_rate": 1.2538814693959294e-05,
+      "loss": 1.9642,
+      "step": 726
+    },
+    {
+      "epoch": 1.727705112960761,
+      "grad_norm": 1.0606410107508442,
+      "learning_rate": 1.2520281635971103e-05,
+      "loss": 1.9007,
+      "step": 727
+    },
+    {
+      "epoch": 1.7300832342449466,
+      "grad_norm": 1.064544615609171,
+      "learning_rate": 1.2501739329702453e-05,
+      "loss": 2.0957,
+      "step": 728
+    },
+    {
+      "epoch": 1.732461355529132,
+      "grad_norm": 1.2317528483018596,
+      "learning_rate": 1.2483187843195128e-05,
+      "loss": 2.0826,
+      "step": 729
+    },
+    {
+      "epoch": 1.7348394768133175,
+      "grad_norm": 0.8079713987023232,
+      "learning_rate": 1.2464627244524595e-05,
+      "loss": 1.9538,
+      "step": 730
+    },
+    {
+      "epoch": 1.7372175980975029,
+      "grad_norm": 1.2758113698136755,
+      "learning_rate": 1.2446057601799753e-05,
+      "loss": 1.9805,
+      "step": 731
+    },
+    {
+      "epoch": 1.7395957193816884,
+      "grad_norm": 1.0629931764775897,
+      "learning_rate": 1.2427478983162694e-05,
+      "loss": 1.9351,
+      "step": 732
+    },
+    {
+      "epoch": 1.741973840665874,
+      "grad_norm": 0.9152635150811873,
+      "learning_rate": 1.2408891456788457e-05,
+      "loss": 1.8034,
+      "step": 733
+    },
+    {
+      "epoch": 1.7443519619500596,
+      "grad_norm": 1.1182721783403513,
+      "learning_rate": 1.2390295090884752e-05,
+      "loss": 2.0279,
+      "step": 734
+    },
+    {
+      "epoch": 1.746730083234245,
+      "grad_norm": 0.8598255371847341,
+      "learning_rate": 1.2371689953691733e-05,
+      "loss": 1.8253,
+      "step": 735
+    },
+    {
+      "epoch": 1.746730083234245,
+      "eval_loss": 2.387751340866089,
+      "eval_runtime": 66.4359,
+      "eval_samples_per_second": 9.618,
+      "eval_steps_per_second": 1.204,
+      "step": 735
+    },
+    {
+      "epoch": 1.7491082045184303,
+      "grad_norm": 0.9861227291838077,
+      "learning_rate": 1.2353076113481742e-05,
+      "loss": 1.9038,
+      "step": 736
+    },
+    {
+      "epoch": 1.7514863258026159,
+      "grad_norm": 1.039133223415763,
+      "learning_rate": 1.2334453638559057e-05,
+      "loss": 2.0651,
+      "step": 737
+    },
+    {
+      "epoch": 1.7538644470868014,
+      "grad_norm": 1.036263407591309,
+      "learning_rate": 1.2315822597259636e-05,
+      "loss": 1.7324,
+      "step": 738
+    },
+    {
+      "epoch": 1.756242568370987,
+      "grad_norm": 0.9557591103456263,
+      "learning_rate": 1.2297183057950886e-05,
+      "loss": 2.0171,
+      "step": 739
+    },
+    {
+      "epoch": 1.7586206896551724,
+      "grad_norm": 0.9906419044217084,
+      "learning_rate": 1.2278535089031377e-05,
+      "loss": 1.8361,
+      "step": 740
+    },
+    {
+      "epoch": 1.760998810939358,
+      "grad_norm": 1.0631674141525231,
+      "learning_rate": 1.2259878758930627e-05,
+      "loss": 1.9345,
+      "step": 741
+    },
+    {
+      "epoch": 1.7633769322235433,
+      "grad_norm": 1.005919565429525,
+      "learning_rate": 1.2241214136108834e-05,
+      "loss": 2.0726,
+      "step": 742
+    },
+    {
+      "epoch": 1.7657550535077289,
+      "grad_norm": 0.8171183104600345,
+      "learning_rate": 1.222254128905662e-05,
+      "loss": 1.9147,
+      "step": 743
+    },
+    {
+      "epoch": 1.7681331747919145,
+      "grad_norm": 1.025525655835034,
+      "learning_rate": 1.220386028629479e-05,
+      "loss": 1.9566,
+      "step": 744
+    },
+    {
+      "epoch": 1.7705112960761,
+      "grad_norm": 1.1199177387750694,
+      "learning_rate": 1.218517119637408e-05,
+      "loss": 1.8823,
+      "step": 745
+    },
+    {
+      "epoch": 1.7728894173602854,
+      "grad_norm": 0.8074812474679398,
+      "learning_rate": 1.2166474087874893e-05,
+      "loss": 1.8039,
+      "step": 746
+    },
+    {
+      "epoch": 1.7752675386444707,
+      "grad_norm": 0.8574176753575323,
+      "learning_rate": 1.2147769029407069e-05,
+      "loss": 1.911,
+      "step": 747
+    },
+    {
+      "epoch": 1.7776456599286563,
+      "grad_norm": 1.0390138098655621,
+      "learning_rate": 1.212905608960961e-05,
+      "loss": 1.6973,
+      "step": 748
+    },
+    {
+      "epoch": 1.7800237812128419,
+      "grad_norm": 0.9501509089266569,
+      "learning_rate": 1.2110335337150445e-05,
+      "loss": 2.0167,
+      "step": 749
+    },
+    {
+      "epoch": 1.7824019024970275,
+      "grad_norm": 1.021610077627784,
+      "learning_rate": 1.209160684072617e-05,
+      "loss": 2.032,
+      "step": 750
+    },
+    {
+      "epoch": 1.7847800237812128,
+      "grad_norm": 1.067584580316804,
+      "learning_rate": 1.2072870669061793e-05,
+      "loss": 2.0627,
+      "step": 751
+    },
+    {
+      "epoch": 1.7871581450653984,
+      "grad_norm": 0.971924208763616,
+      "learning_rate": 1.2054126890910499e-05,
+      "loss": 1.9211,
+      "step": 752
+    },
+    {
+      "epoch": 1.7895362663495837,
+      "grad_norm": 1.059861690139045,
+      "learning_rate": 1.2035375575053373e-05,
+      "loss": 1.8785,
+      "step": 753
+    },
+    {
+      "epoch": 1.7919143876337693,
+      "grad_norm": 0.9537414827635367,
+      "learning_rate": 1.2016616790299165e-05,
+      "loss": 1.9083,
+      "step": 754
+    },
+    {
+      "epoch": 1.7942925089179549,
+      "grad_norm": 0.9591859502901945,
+      "learning_rate": 1.1997850605484034e-05,
+      "loss": 2.0818,
+      "step": 755
+    },
+    {
+      "epoch": 1.7966706302021405,
+      "grad_norm": 1.0432501587679868,
+      "learning_rate": 1.1979077089471288e-05,
+      "loss": 1.9283,
+      "step": 756
+    },
+    {
+      "epoch": 1.7990487514863258,
+      "grad_norm": 0.9516347204385417,
+      "learning_rate": 1.1960296311151143e-05,
+      "loss": 2.0293,
+      "step": 757
+    },
+    {
+      "epoch": 1.8014268727705112,
+      "grad_norm": 1.0424747053382684,
+      "learning_rate": 1.1941508339440467e-05,
+      "loss": 2.0464,
+      "step": 758
+    },
+    {
+      "epoch": 1.8038049940546967,
+      "grad_norm": 1.0574987855006235,
+      "learning_rate": 1.1922713243282514e-05,
+      "loss": 2.0236,
+      "step": 759
+    },
+    {
+      "epoch": 1.8061831153388823,
+      "grad_norm": 1.0377956706796805,
+      "learning_rate": 1.1903911091646684e-05,
+      "loss": 1.9934,
+      "step": 760
+    },
+    {
+      "epoch": 1.808561236623068,
+      "grad_norm": 1.0769312593664324,
+      "learning_rate": 1.1885101953528279e-05,
+      "loss": 2.1783,
+      "step": 761
+    },
+    {
+      "epoch": 1.8109393579072532,
+      "grad_norm": 0.9429126781570604,
+      "learning_rate": 1.1866285897948227e-05,
+      "loss": 1.899,
+      "step": 762
+    },
+    {
+      "epoch": 1.8133174791914386,
+      "grad_norm": 1.080323404458619,
+      "learning_rate": 1.1847462993952842e-05,
+      "loss": 1.9386,
+      "step": 763
+    },
+    {
+      "epoch": 1.8156956004756242,
+      "grad_norm": 0.9840126348452174,
+      "learning_rate": 1.1828633310613569e-05,
+      "loss": 2.0926,
+      "step": 764
+    },
+    {
+      "epoch": 1.8180737217598097,
+      "grad_norm": 1.1925367572998558,
+      "learning_rate": 1.180979691702673e-05,
+      "loss": 1.8951,
+      "step": 765
+    },
+    {
+      "epoch": 1.8204518430439953,
+      "grad_norm": 0.9952376958926185,
+      "learning_rate": 1.1790953882313271e-05,
+      "loss": 1.9651,
+      "step": 766
+    },
+    {
+      "epoch": 1.822829964328181,
+      "grad_norm": 0.936960021257687,
+      "learning_rate": 1.1772104275618512e-05,
+      "loss": 1.9273,
+      "step": 767
+    },
+    {
+      "epoch": 1.8252080856123662,
+      "grad_norm": 0.9274559160209536,
+      "learning_rate": 1.175324816611188e-05,
+      "loss": 1.9401,
+      "step": 768
+    },
+    {
+      "epoch": 1.8275862068965516,
+      "grad_norm": 1.0137182594497265,
+      "learning_rate": 1.1734385622986668e-05,
+      "loss": 2.0039,
+      "step": 769
+    },
+    {
+      "epoch": 1.8299643281807372,
+      "grad_norm": 1.162702633094251,
+      "learning_rate": 1.1715516715459784e-05,
+      "loss": 1.993,
+      "step": 770
+    },
+    {
+      "epoch": 1.8323424494649228,
+      "grad_norm": 0.9064383127278726,
+      "learning_rate": 1.1696641512771482e-05,
+      "loss": 2.0824,
+      "step": 771
+    },
+    {
+      "epoch": 1.8347205707491083,
+      "grad_norm": 0.9662577578484263,
+      "learning_rate": 1.1677760084185123e-05,
+      "loss": 1.7069,
+      "step": 772
+    },
+    {
+      "epoch": 1.8370986920332937,
+      "grad_norm": 0.8709375549819791,
+      "learning_rate": 1.1658872498986905e-05,
+      "loss": 2.0223,
+      "step": 773
+    },
+    {
+      "epoch": 1.839476813317479,
+      "grad_norm": 1.145853044079966,
+      "learning_rate": 1.1639978826485629e-05,
+      "loss": 1.9748,
+      "step": 774
+    },
+    {
+      "epoch": 1.8418549346016646,
+      "grad_norm": 1.1254237494407062,
+      "learning_rate": 1.1621079136012426e-05,
+      "loss": 1.9803,
+      "step": 775
+    },
+    {
+      "epoch": 1.8442330558858502,
+      "grad_norm": 1.10543629656252,
+      "learning_rate": 1.160217349692051e-05,
+      "loss": 1.9952,
+      "step": 776
+    },
+    {
+      "epoch": 1.8466111771700358,
+      "grad_norm": 1.0421494399227835,
+      "learning_rate": 1.1583261978584934e-05,
+      "loss": 2.2644,
+      "step": 777
+    },
+    {
+      "epoch": 1.8489892984542213,
+      "grad_norm": 1.157252899286374,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 1.8651,
+      "step": 778
+    },
+    {
+      "epoch": 1.8513674197384067,
+      "grad_norm": 1.1787150544248581,
+      "learning_rate": 1.1545421581790579e-05,
+      "loss": 2.006,
+      "step": 779
+    },
+    {
+      "epoch": 1.853745541022592,
+      "grad_norm": 0.9108745110078712,
+      "learning_rate": 1.1526492842188746e-05,
+      "loss": 1.9716,
+      "step": 780
+    },
+    {
+      "epoch": 1.8561236623067776,
+      "grad_norm": 1.0584595603639568,
+      "learning_rate": 1.1507558501056622e-05,
+      "loss": 2.0011,
+      "step": 781
+    },
+    {
+      "epoch": 1.8585017835909632,
+      "grad_norm": 1.0921406194015688,
+      "learning_rate": 1.1488618627874581e-05,
+      "loss": 1.8896,
+      "step": 782
+    },
+    {
+      "epoch": 1.8608799048751488,
+      "grad_norm": 1.0068894499355592,
+      "learning_rate": 1.1469673292143284e-05,
+      "loss": 1.9903,
+      "step": 783
+    },
+    {
+      "epoch": 1.8632580261593341,
+      "grad_norm": 1.0157407453051925,
+      "learning_rate": 1.145072256338345e-05,
+      "loss": 1.8454,
+      "step": 784
+    },
+    {
+      "epoch": 1.8656361474435195,
+      "grad_norm": 1.0130783012738553,
+      "learning_rate": 1.1431766511135581e-05,
+      "loss": 1.856,
+      "step": 785
+    },
+    {
+      "epoch": 1.868014268727705,
+      "grad_norm": 1.1047664955736678,
+      "learning_rate": 1.1412805204959719e-05,
+      "loss": 1.881,
+      "step": 786
+    },
+    {
+      "epoch": 1.8703923900118906,
+      "grad_norm": 0.9130130901731418,
+      "learning_rate": 1.1393838714435176e-05,
+      "loss": 1.9902,
+      "step": 787
+    },
+    {
+      "epoch": 1.8727705112960762,
+      "grad_norm": 0.8444730119781477,
+      "learning_rate": 1.1374867109160295e-05,
+      "loss": 1.8033,
+      "step": 788
+    },
+    {
+      "epoch": 1.8751486325802615,
+      "grad_norm": 1.0068775645787829,
+      "learning_rate": 1.135589045875219e-05,
+      "loss": 1.8523,
+      "step": 789
+    },
+    {
+      "epoch": 1.8775267538644471,
+      "grad_norm": 1.0137172293384147,
+      "learning_rate": 1.1336908832846485e-05,
+      "loss": 1.8672,
+      "step": 790
+    },
+    {
+      "epoch": 1.8799048751486325,
+      "grad_norm": 1.1230734224665218,
+      "learning_rate": 1.1317922301097064e-05,
+      "loss": 1.9669,
+      "step": 791
+    },
+    {
+      "epoch": 1.882282996432818,
+      "grad_norm": 0.8743774773477322,
+      "learning_rate": 1.1298930933175805e-05,
+      "loss": 1.7966,
+      "step": 792
+    },
+    {
+      "epoch": 1.8846611177170036,
+      "grad_norm": 1.321843260348692,
+      "learning_rate": 1.1279934798772344e-05,
+      "loss": 1.9052,
+      "step": 793
+    },
+    {
+      "epoch": 1.8870392390011892,
+      "grad_norm": 0.9265508467848211,
+      "learning_rate": 1.12609339675938e-05,
+      "loss": 2.0626,
+      "step": 794
+    },
+    {
+      "epoch": 1.8894173602853745,
+      "grad_norm": 1.065099635395474,
+      "learning_rate": 1.1241928509364533e-05,
+      "loss": 1.8472,
+      "step": 795
+    },
+    {
+      "epoch": 1.89179548156956,
+      "grad_norm": 0.9336444235822928,
+      "learning_rate": 1.1222918493825876e-05,
+      "loss": 2.0588,
+      "step": 796
+    },
+    {
+      "epoch": 1.8941736028537455,
+      "grad_norm": 0.9901340362445608,
+      "learning_rate": 1.1203903990735888e-05,
+      "loss": 1.969,
+      "step": 797
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 1.1007028104085261,
+      "learning_rate": 1.1184885069869096e-05,
+      "loss": 1.8668,
+      "step": 798
+    },
+    {
+      "epoch": 1.8989298454221166,
+      "grad_norm": 0.8704786291588872,
+      "learning_rate": 1.1165861801016235e-05,
+      "loss": 1.7756,
+      "step": 799
+    },
+    {
+      "epoch": 1.901307966706302,
+      "grad_norm": 0.9586858624614886,
+      "learning_rate": 1.1146834253984008e-05,
+      "loss": 2.1082,
+      "step": 800
+    },
+    {
+      "epoch": 1.9036860879904876,
+      "grad_norm": 0.9935316309566465,
+      "learning_rate": 1.1127802498594792e-05,
+      "loss": 2.0898,
+      "step": 801
+    },
+    {
+      "epoch": 1.906064209274673,
+      "grad_norm": 1.0128018872894315,
+      "learning_rate": 1.110876660468643e-05,
+      "loss": 2.0126,
+      "step": 802
+    },
+    {
+      "epoch": 1.9084423305588585,
+      "grad_norm": 1.0425566210644994,
+      "learning_rate": 1.108972664211194e-05,
+      "loss": 1.8575,
+      "step": 803
+    },
+    {
+      "epoch": 1.910820451843044,
+      "grad_norm": 1.155933364218014,
+      "learning_rate": 1.1070682680739275e-05,
+      "loss": 1.9965,
+      "step": 804
+    },
+    {
+      "epoch": 1.9131985731272296,
+      "grad_norm": 1.4109307764012204,
+      "learning_rate": 1.105163479045106e-05,
+      "loss": 1.7907,
+      "step": 805
+    },
+    {
+      "epoch": 1.915576694411415,
+      "grad_norm": 1.019545311598646,
+      "learning_rate": 1.1032583041144334e-05,
+      "loss": 1.7445,
+      "step": 806
+    },
+    {
+      "epoch": 1.9179548156956003,
+      "grad_norm": 0.8852176465494631,
+      "learning_rate": 1.1013527502730301e-05,
+      "loss": 1.8363,
+      "step": 807
+    },
+    {
+      "epoch": 1.920332936979786,
+      "grad_norm": 0.967544348388286,
+      "learning_rate": 1.0994468245134071e-05,
+      "loss": 1.7991,
+      "step": 808
+    },
+    {
+      "epoch": 1.9227110582639715,
+      "grad_norm": 1.005400241628268,
+      "learning_rate": 1.0975405338294399e-05,
+      "loss": 1.8783,
+      "step": 809
+    },
+    {
+      "epoch": 1.925089179548157,
+      "grad_norm": 0.8762088510461341,
+      "learning_rate": 1.0956338852163424e-05,
+      "loss": 1.8432,
+      "step": 810
+    },
+    {
+      "epoch": 1.9274673008323424,
+      "grad_norm": 1.0033258791075268,
+      "learning_rate": 1.0937268856706435e-05,
+      "loss": 1.9274,
+      "step": 811
+    },
+    {
+      "epoch": 1.929845422116528,
+      "grad_norm": 0.9911088707682482,
+      "learning_rate": 1.0918195421901583e-05,
+      "loss": 1.7392,
+      "step": 812
+    },
+    {
+      "epoch": 1.9322235434007133,
+      "grad_norm": 1.2905322005763973,
+      "learning_rate": 1.089911861773965e-05,
+      "loss": 2.1347,
+      "step": 813
+    },
+    {
+      "epoch": 1.934601664684899,
+      "grad_norm": 1.014836431956825,
+      "learning_rate": 1.0880038514223778e-05,
+      "loss": 1.9524,
+      "step": 814
+    },
+    {
+      "epoch": 1.9369797859690845,
+      "grad_norm": 0.9524664386824049,
+      "learning_rate": 1.0860955181369219e-05,
+      "loss": 1.9656,
+      "step": 815
+    },
+    {
+      "epoch": 1.93935790725327,
+      "grad_norm": 1.1264640139375,
+      "learning_rate": 1.0841868689203072e-05,
+      "loss": 1.8089,
+      "step": 816
+    },
+    {
+      "epoch": 1.9417360285374554,
+      "grad_norm": 0.8635336405252807,
+      "learning_rate": 1.0822779107764028e-05,
+      "loss": 1.908,
+      "step": 817
+    },
+    {
+      "epoch": 1.9441141498216408,
+      "grad_norm": 1.0677562500326228,
+      "learning_rate": 1.0803686507102118e-05,
+      "loss": 1.9362,
+      "step": 818
+    },
+    {
+      "epoch": 1.9464922711058263,
+      "grad_norm": 1.0413287807888576,
+      "learning_rate": 1.0784590957278452e-05,
+      "loss": 1.8287,
+      "step": 819
+    },
+    {
+      "epoch": 1.948870392390012,
+      "grad_norm": 1.1002052999274758,
+      "learning_rate": 1.076549252836496e-05,
+      "loss": 2.0404,
+      "step": 820
+    },
+    {
+      "epoch": 1.9512485136741975,
+      "grad_norm": 1.0659689575025673,
+      "learning_rate": 1.0746391290444136e-05,
+      "loss": 2.0779,
+      "step": 821
+    },
+    {
+      "epoch": 1.9536266349583828,
+      "grad_norm": 1.0634086392515605,
+      "learning_rate": 1.0727287313608783e-05,
+      "loss": 1.9018,
+      "step": 822
+    },
+    {
+      "epoch": 1.9560047562425684,
+      "grad_norm": 0.8849872440417224,
+      "learning_rate": 1.0708180667961758e-05,
+      "loss": 2.1171,
+      "step": 823
+    },
+    {
+      "epoch": 1.9583828775267538,
+      "grad_norm": 1.0283232728798035,
+      "learning_rate": 1.0689071423615708e-05,
+      "loss": 1.8658,
+      "step": 824
+    },
+    {
+      "epoch": 1.9607609988109393,
+      "grad_norm": 1.1823895638000625,
+      "learning_rate": 1.0669959650692818e-05,
+      "loss": 1.6882,
+      "step": 825
+    },
+    {
+      "epoch": 1.963139120095125,
+      "grad_norm": 1.1175568913711953,
+      "learning_rate": 1.0650845419324544e-05,
+      "loss": 1.855,
+      "step": 826
+    },
+    {
+      "epoch": 1.9655172413793105,
+      "grad_norm": 1.1061515815620258,
+      "learning_rate": 1.0631728799651374e-05,
+      "loss": 2.0233,
+      "step": 827
+    },
+    {
+      "epoch": 1.9678953626634959,
+      "grad_norm": 0.9330753316753481,
+      "learning_rate": 1.0612609861822556e-05,
+      "loss": 2.0104,
+      "step": 828
+    },
+    {
+      "epoch": 1.9702734839476812,
+      "grad_norm": 1.0424417166124378,
+      "learning_rate": 1.059348867599584e-05,
+      "loss": 1.9829,
+      "step": 829
+    },
+    {
+      "epoch": 1.9726516052318668,
+      "grad_norm": 0.9251278544848563,
+      "learning_rate": 1.0574365312337235e-05,
+      "loss": 1.6856,
+      "step": 830
+    },
+    {
+      "epoch": 1.9750297265160524,
+      "grad_norm": 1.0217360426495539,
+      "learning_rate": 1.055523984102073e-05,
+      "loss": 1.9015,
+      "step": 831
+    },
+    {
+      "epoch": 1.977407847800238,
+      "grad_norm": 1.01682991161189,
+      "learning_rate": 1.0536112332228057e-05,
+      "loss": 1.917,
+      "step": 832
+    },
+    {
+      "epoch": 1.9797859690844233,
+      "grad_norm": 0.9565389300837592,
+      "learning_rate": 1.0516982856148421e-05,
+      "loss": 1.8118,
+      "step": 833
+    },
+    {
+      "epoch": 1.9821640903686086,
+      "grad_norm": 1.104212144557824,
+      "learning_rate": 1.0497851482978247e-05,
+      "loss": 1.8446,
+      "step": 834
+    },
+    {
+      "epoch": 1.9845422116527942,
+      "grad_norm": 0.9911600279737486,
+      "learning_rate": 1.0478718282920922e-05,
+      "loss": 1.9056,
+      "step": 835
+    },
+    {
+      "epoch": 1.9869203329369798,
+      "grad_norm": 0.9981165586294449,
+      "learning_rate": 1.0459583326186532e-05,
+      "loss": 2.019,
+      "step": 836
+    },
+    {
+      "epoch": 1.9892984542211654,
+      "grad_norm": 1.2857901353617724,
+      "learning_rate": 1.0440446682991617e-05,
+      "loss": 1.8977,
+      "step": 837
+    },
+    {
+      "epoch": 1.991676575505351,
+      "grad_norm": 1.2212775483605185,
+      "learning_rate": 1.0421308423558898e-05,
+      "loss": 1.9891,
+      "step": 838
+    },
+    {
+      "epoch": 1.9940546967895363,
+      "grad_norm": 1.0720928120882365,
+      "learning_rate": 1.0402168618117038e-05,
+      "loss": 2.1227,
+      "step": 839
+    },
+    {
+      "epoch": 1.9964328180737216,
+      "grad_norm": 1.0906773289299942,
+      "learning_rate": 1.0383027336900356e-05,
+      "loss": 1.9278,
+      "step": 840
+    },
+    {
+      "epoch": 1.9964328180737216,
+      "eval_loss": 2.3622491359710693,
+      "eval_runtime": 65.5731,
+      "eval_samples_per_second": 9.745,
+      "eval_steps_per_second": 1.22,
+      "step": 840
+    },
+    {
+      "epoch": 1.9988109393579072,
+      "grad_norm": 0.9597175384609758,
+      "learning_rate": 1.0363884650148601e-05,
+      "loss": 2.1366,
+      "step": 841
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6494060559715102,
+      "learning_rate": 1.0344740628106673e-05,
+      "loss": 2.0977,
+      "step": 842
+    },
+    {
+      "epoch": 2.0023781212841856,
+      "grad_norm": 1.8572000704827933,
+      "learning_rate": 1.0325595341024377e-05,
+      "loss": 1.804,
+      "step": 843
+    },
+    {
+      "epoch": 2.004756242568371,
+      "grad_norm": 1.6109351938216152,
+      "learning_rate": 1.0306448859156155e-05,
+      "loss": 1.6572,
+      "step": 844
+    },
+    {
+      "epoch": 2.0071343638525563,
+      "grad_norm": 1.682163858699801,
+      "learning_rate": 1.0287301252760833e-05,
+      "loss": 1.9179,
+      "step": 845
+    },
+    {
+      "epoch": 2.009512485136742,
+      "grad_norm": 1.4642932185295525,
+      "learning_rate": 1.0268152592101366e-05,
+      "loss": 1.7944,
+      "step": 846
+    },
+    {
+      "epoch": 2.0118906064209274,
+      "grad_norm": 2.298918914235818,
+      "learning_rate": 1.024900294744458e-05,
+      "loss": 1.4883,
+      "step": 847
+    },
+    {
+      "epoch": 2.014268727705113,
+      "grad_norm": 2.1569642060668723,
+      "learning_rate": 1.0229852389060905e-05,
+      "loss": 1.7754,
+      "step": 848
+    },
+    {
+      "epoch": 2.0166468489892986,
+      "grad_norm": 1.650877681094919,
+      "learning_rate": 1.0210700987224134e-05,
+      "loss": 1.7676,
+      "step": 849
+    },
+    {
+      "epoch": 2.019024970273484,
+      "grad_norm": 1.4229362966156138,
+      "learning_rate": 1.0191548812211143e-05,
+      "loss": 1.6903,
+      "step": 850
+    },
+    {
+      "epoch": 2.0214030915576693,
+      "grad_norm": 1.3427234695472385,
+      "learning_rate": 1.0172395934301653e-05,
+      "loss": 1.8036,
+      "step": 851
+    },
+    {
+      "epoch": 2.023781212841855,
+      "grad_norm": 1.4370215727780862,
+      "learning_rate": 1.0153242423777964e-05,
+      "loss": 1.6991,
+      "step": 852
+    },
+    {
+      "epoch": 2.0261593341260404,
+      "grad_norm": 1.2143026318203651,
+      "learning_rate": 1.0134088350924699e-05,
+      "loss": 1.7004,
+      "step": 853
+    },
+    {
+      "epoch": 2.028537455410226,
+      "grad_norm": 1.2634866035152414,
+      "learning_rate": 1.0114933786028534e-05,
+      "loss": 1.7919,
+      "step": 854
+    },
+    {
+      "epoch": 2.0309155766944116,
+      "grad_norm": 1.3125463113477043,
+      "learning_rate": 1.009577879937796e-05,
+      "loss": 1.8493,
+      "step": 855
+    },
+    {
+      "epoch": 2.0332936979785967,
+      "grad_norm": 1.1854528046002857,
+      "learning_rate": 1.0076623461263017e-05,
+      "loss": 1.7386,
+      "step": 856
+    },
+    {
+      "epoch": 2.0356718192627823,
+      "grad_norm": 1.2742920280289385,
+      "learning_rate": 1.005746784197503e-05,
+      "loss": 1.5267,
+      "step": 857
+    },
+    {
+      "epoch": 2.038049940546968,
+      "grad_norm": 1.4794216244383256,
+      "learning_rate": 1.003831201180636e-05,
+      "loss": 1.68,
+      "step": 858
+    },
+    {
+      "epoch": 2.0404280618311534,
+      "grad_norm": 1.1043401807863704,
+      "learning_rate": 1.0019156041050134e-05,
+      "loss": 1.7469,
+      "step": 859
+    },
+    {
+      "epoch": 2.042806183115339,
+      "grad_norm": 1.3043477136558799,
+      "learning_rate": 1e-05,
+      "loss": 1.7495,
+      "step": 860
+    },
+    {
+      "epoch": 2.0451843043995246,
+      "grad_norm": 1.1439412634195671,
+      "learning_rate": 9.98084395894987e-06,
+      "loss": 1.7956,
+      "step": 861
+    },
+    {
+      "epoch": 2.0475624256837097,
+      "grad_norm": 1.1847513743650282,
+      "learning_rate": 9.961687988193642e-06,
+      "loss": 1.6905,
+      "step": 862
+    },
+    {
+      "epoch": 2.0499405469678953,
+      "grad_norm": 1.230107373003021,
+      "learning_rate": 9.942532158024971e-06,
+      "loss": 1.637,
+      "step": 863
+    },
+    {
+      "epoch": 2.052318668252081,
+      "grad_norm": 1.331643711464931,
+      "learning_rate": 9.923376538736985e-06,
+      "loss": 1.9403,
+      "step": 864
+    },
+    {
+      "epoch": 2.0546967895362664,
+      "grad_norm": 1.3885388021890999,
+      "learning_rate": 9.904221200622042e-06,
+      "loss": 1.7906,
+      "step": 865
+    },
+    {
+      "epoch": 2.057074910820452,
+      "grad_norm": 1.204006509857973,
+      "learning_rate": 9.885066213971471e-06,
+      "loss": 1.5899,
+      "step": 866
+    },
+    {
+      "epoch": 2.059453032104637,
+      "grad_norm": 1.1475843090693567,
+      "learning_rate": 9.865911649075305e-06,
+      "loss": 1.7476,
+      "step": 867
+    },
+    {
+      "epoch": 2.0618311533888227,
+      "grad_norm": 1.5532196494457342,
+      "learning_rate": 9.846757576222038e-06,
+      "loss": 2.0227,
+      "step": 868
+    },
+    {
+      "epoch": 2.0642092746730083,
+      "grad_norm": 1.1040634362125452,
+      "learning_rate": 9.82760406569835e-06,
+      "loss": 1.7182,
+      "step": 869
+    },
+    {
+      "epoch": 2.066587395957194,
+      "grad_norm": 1.2299107164007779,
+      "learning_rate": 9.80845118778886e-06,
+      "loss": 1.8007,
+      "step": 870
+    },
+    {
+      "epoch": 2.0689655172413794,
+      "grad_norm": 1.3188262031162163,
+      "learning_rate": 9.78929901277587e-06,
+      "loss": 1.8093,
+      "step": 871
+    },
+    {
+      "epoch": 2.0713436385255646,
+      "grad_norm": 1.303921293625063,
+      "learning_rate": 9.770147610939098e-06,
+      "loss": 1.3981,
+      "step": 872
+    },
+    {
+      "epoch": 2.07372175980975,
+      "grad_norm": 1.1390122295999698,
+      "learning_rate": 9.750997052555423e-06,
+      "loss": 1.6957,
+      "step": 873
+    },
+    {
+      "epoch": 2.0760998810939357,
+      "grad_norm": 1.7116484489773238,
+      "learning_rate": 9.731847407898637e-06,
+      "loss": 1.7392,
+      "step": 874
+    },
+    {
+      "epoch": 2.0784780023781213,
+      "grad_norm": 1.2294423692022918,
+      "learning_rate": 9.712698747239172e-06,
+      "loss": 1.7854,
+      "step": 875
+    },
+    {
+      "epoch": 2.080856123662307,
+      "grad_norm": 1.2941804638097063,
+      "learning_rate": 9.693551140843848e-06,
+      "loss": 1.6763,
+      "step": 876
+    },
+    {
+      "epoch": 2.0832342449464925,
+      "grad_norm": 1.167524044387485,
+      "learning_rate": 9.674404658975627e-06,
+      "loss": 1.7381,
+      "step": 877
+    },
+    {
+      "epoch": 2.0856123662306776,
+      "grad_norm": 1.3380153711442146,
+      "learning_rate": 9.655259371893329e-06,
+      "loss": 1.7105,
+      "step": 878
+    },
+    {
+      "epoch": 2.087990487514863,
+      "grad_norm": 1.3080896822024415,
+      "learning_rate": 9.636115349851402e-06,
+      "loss": 1.7833,
+      "step": 879
+    },
+    {
+      "epoch": 2.0903686087990487,
+      "grad_norm": 1.2366799144892753,
+      "learning_rate": 9.616972663099648e-06,
+      "loss": 1.679,
+      "step": 880
+    },
+    {
+      "epoch": 2.0927467300832343,
+      "grad_norm": 1.1601075579319895,
+      "learning_rate": 9.597831381882966e-06,
+      "loss": 1.6583,
+      "step": 881
+    },
+    {
+      "epoch": 2.09512485136742,
+      "grad_norm": 1.2534594907956864,
+      "learning_rate": 9.578691576441104e-06,
+      "loss": 1.4498,
+      "step": 882
+    },
+    {
+      "epoch": 2.097502972651605,
+      "grad_norm": 1.3319983592609854,
+      "learning_rate": 9.559553317008386e-06,
+      "loss": 1.791,
+      "step": 883
+    },
+    {
+      "epoch": 2.0998810939357906,
+      "grad_norm": 1.1817544740271848,
+      "learning_rate": 9.540416673813471e-06,
+      "loss": 1.6665,
+      "step": 884
+    },
+    {
+      "epoch": 2.102259215219976,
+      "grad_norm": 1.3535125184572838,
+      "learning_rate": 9.521281717079082e-06,
+      "loss": 1.6581,
+      "step": 885
+    },
+    {
+      "epoch": 2.1046373365041617,
+      "grad_norm": 1.3813623473085552,
+      "learning_rate": 9.502148517021757e-06,
+      "loss": 1.9743,
+      "step": 886
+    },
+    {
+      "epoch": 2.1070154577883473,
+      "grad_norm": 1.1982300911541166,
+      "learning_rate": 9.483017143851582e-06,
+      "loss": 1.6314,
+      "step": 887
+    },
+    {
+      "epoch": 2.109393579072533,
+      "grad_norm": 1.270544773734369,
+      "learning_rate": 9.463887667771946e-06,
+      "loss": 1.559,
+      "step": 888
+    },
+    {
+      "epoch": 2.111771700356718,
+      "grad_norm": 1.1811810082167495,
+      "learning_rate": 9.444760158979272e-06,
+      "loss": 1.6682,
+      "step": 889
+    },
+    {
+      "epoch": 2.1141498216409036,
+      "grad_norm": 1.1438900433747767,
+      "learning_rate": 9.425634687662768e-06,
+      "loss": 1.6899,
+      "step": 890
+    },
+    {
+      "epoch": 2.116527942925089,
+      "grad_norm": 1.2892989413709668,
+      "learning_rate": 9.406511324004162e-06,
+      "loss": 1.7346,
+      "step": 891
+    },
+    {
+      "epoch": 2.1189060642092747,
+      "grad_norm": 1.1712789054414114,
+      "learning_rate": 9.387390138177447e-06,
+      "loss": 1.6321,
+      "step": 892
+    },
+    {
+      "epoch": 2.1212841854934603,
+      "grad_norm": 1.1012248650637824,
+      "learning_rate": 9.368271200348627e-06,
+      "loss": 1.5844,
+      "step": 893
+    },
+    {
+      "epoch": 2.1236623067776454,
+      "grad_norm": 1.2478720681495163,
+      "learning_rate": 9.349154580675457e-06,
+      "loss": 1.4902,
+      "step": 894
+    },
+    {
+      "epoch": 2.126040428061831,
+      "grad_norm": 1.268299866629765,
+      "learning_rate": 9.330040349307185e-06,
+      "loss": 1.7093,
+      "step": 895
+    },
+    {
+      "epoch": 2.1284185493460166,
+      "grad_norm": 1.2311038089588855,
+      "learning_rate": 9.310928576384293e-06,
+      "loss": 1.6951,
+      "step": 896
+    },
+    {
+      "epoch": 2.130796670630202,
+      "grad_norm": 1.3174407079124912,
+      "learning_rate": 9.291819332038243e-06,
+      "loss": 1.8443,
+      "step": 897
+    },
+    {
+      "epoch": 2.1331747919143877,
+      "grad_norm": 1.2906831927984828,
+      "learning_rate": 9.272712686391219e-06,
+      "loss": 1.8221,
+      "step": 898
+    },
+    {
+      "epoch": 2.1355529131985733,
+      "grad_norm": 1.2522791578826056,
+      "learning_rate": 9.253608709555869e-06,
+      "loss": 1.6128,
+      "step": 899
+    },
+    {
+      "epoch": 2.1379310344827585,
+      "grad_norm": 1.0934517484029729,
+      "learning_rate": 9.234507471635043e-06,
+      "loss": 1.6859,
+      "step": 900
+    },
+    {
+      "epoch": 2.140309155766944,
+      "grad_norm": 1.194959938259093,
+      "learning_rate": 9.215409042721553e-06,
+      "loss": 1.7118,
+      "step": 901
+    },
+    {
+      "epoch": 2.1426872770511296,
+      "grad_norm": 1.1113101873824878,
+      "learning_rate": 9.196313492897885e-06,
+      "loss": 1.7543,
+      "step": 902
+    },
+    {
+      "epoch": 2.145065398335315,
+      "grad_norm": 1.207090436132185,
+      "learning_rate": 9.177220892235976e-06,
+      "loss": 1.8335,
+      "step": 903
+    },
+    {
+      "epoch": 2.1474435196195008,
+      "grad_norm": 1.1323648129867685,
+      "learning_rate": 9.15813131079693e-06,
+      "loss": 1.622,
+      "step": 904
+    },
+    {
+      "epoch": 2.149821640903686,
+      "grad_norm": 1.1631703384520493,
+      "learning_rate": 9.139044818630784e-06,
+      "loss": 1.734,
+      "step": 905
+    },
+    {
+      "epoch": 2.1521997621878715,
+      "grad_norm": 1.1629589692582991,
+      "learning_rate": 9.119961485776223e-06,
+      "loss": 1.6643,
+      "step": 906
+    },
+    {
+      "epoch": 2.154577883472057,
+      "grad_norm": 1.1621837317471908,
+      "learning_rate": 9.100881382260353e-06,
+      "loss": 1.6431,
+      "step": 907
+    },
+    {
+      "epoch": 2.1569560047562426,
+      "grad_norm": 1.1112908601843028,
+      "learning_rate": 9.08180457809842e-06,
+      "loss": 1.7269,
+      "step": 908
+    },
+    {
+      "epoch": 2.159334126040428,
+      "grad_norm": 1.2755784245054267,
+      "learning_rate": 9.062731143293569e-06,
+      "loss": 1.846,
+      "step": 909
+    },
+    {
+      "epoch": 2.1617122473246138,
+      "grad_norm": 1.1803023595907047,
+      "learning_rate": 9.043661147836578e-06,
+      "loss": 1.6091,
+      "step": 910
+    },
+    {
+      "epoch": 2.164090368608799,
+      "grad_norm": 1.185874011924432,
+      "learning_rate": 9.024594661705605e-06,
+      "loss": 1.6543,
+      "step": 911
+    },
+    {
+      "epoch": 2.1664684898929845,
+      "grad_norm": 1.3961342001389099,
+      "learning_rate": 9.005531754865929e-06,
+      "loss": 1.7178,
+      "step": 912
+    },
+    {
+      "epoch": 2.16884661117717,
+      "grad_norm": 1.333944871158627,
+      "learning_rate": 8.986472497269699e-06,
+      "loss": 1.701,
+      "step": 913
+    },
+    {
+      "epoch": 2.1712247324613556,
+      "grad_norm": 1.6434825073753228,
+      "learning_rate": 8.967416958855666e-06,
+      "loss": 1.8314,
+      "step": 914
+    },
+    {
+      "epoch": 2.173602853745541,
+      "grad_norm": 1.270123334452636,
+      "learning_rate": 8.948365209548943e-06,
+      "loss": 1.8827,
+      "step": 915
+    },
+    {
+      "epoch": 2.1759809750297263,
+      "grad_norm": 1.153010176885914,
+      "learning_rate": 8.929317319260727e-06,
+      "loss": 1.8395,
+      "step": 916
+    },
+    {
+      "epoch": 2.178359096313912,
+      "grad_norm": 1.4178416006595018,
+      "learning_rate": 8.91027335788806e-06,
+      "loss": 1.8333,
+      "step": 917
+    },
+    {
+      "epoch": 2.1807372175980975,
+      "grad_norm": 1.1553451804083652,
+      "learning_rate": 8.891233395313572e-06,
+      "loss": 1.7945,
+      "step": 918
+    },
+    {
+      "epoch": 2.183115338882283,
+      "grad_norm": 1.2293490645752978,
+      "learning_rate": 8.872197501405208e-06,
+      "loss": 1.7454,
+      "step": 919
+    },
+    {
+      "epoch": 2.1854934601664686,
+      "grad_norm": 1.2419979870389453,
+      "learning_rate": 8.853165746015997e-06,
+      "loss": 1.8377,
+      "step": 920
+    },
+    {
+      "epoch": 2.187871581450654,
+      "grad_norm": 1.7147597799918568,
+      "learning_rate": 8.834138198983763e-06,
+      "loss": 1.7252,
+      "step": 921
+    },
+    {
+      "epoch": 2.1902497027348393,
+      "grad_norm": 1.2306648279680026,
+      "learning_rate": 8.815114930130906e-06,
+      "loss": 1.6057,
+      "step": 922
+    },
+    {
+      "epoch": 2.192627824019025,
+      "grad_norm": 1.1859169425356446,
+      "learning_rate": 8.796096009264114e-06,
+      "loss": 1.6553,
+      "step": 923
+    },
+    {
+      "epoch": 2.1950059453032105,
+      "grad_norm": 1.1602261640700344,
+      "learning_rate": 8.777081506174127e-06,
+      "loss": 1.481,
+      "step": 924
+    },
+    {
+      "epoch": 2.197384066587396,
+      "grad_norm": 1.402945789419062,
+      "learning_rate": 8.758071490635468e-06,
+      "loss": 1.9423,
+      "step": 925
+    },
+    {
+      "epoch": 2.1997621878715816,
+      "grad_norm": 1.081930558901166,
+      "learning_rate": 8.739066032406201e-06,
+      "loss": 1.5998,
+      "step": 926
+    },
+    {
+      "epoch": 2.2021403091557668,
+      "grad_norm": 1.1850370906254002,
+      "learning_rate": 8.720065201227656e-06,
+      "loss": 1.4878,
+      "step": 927
+    },
+    {
+      "epoch": 2.2045184304399523,
+      "grad_norm": 1.1307191491979665,
+      "learning_rate": 8.701069066824195e-06,
+      "loss": 1.6825,
+      "step": 928
+    },
+    {
+      "epoch": 2.206896551724138,
+      "grad_norm": 1.2308894920595725,
+      "learning_rate": 8.68207769890294e-06,
+      "loss": 1.6438,
+      "step": 929
+    },
+    {
+      "epoch": 2.2092746730083235,
+      "grad_norm": 1.3885937358394551,
+      "learning_rate": 8.663091167153516e-06,
+      "loss": 1.7407,
+      "step": 930
+    },
+    {
+      "epoch": 2.211652794292509,
+      "grad_norm": 1.3493954256475922,
+      "learning_rate": 8.644109541247811e-06,
+      "loss": 1.7611,
+      "step": 931
+    },
+    {
+      "epoch": 2.2140309155766946,
+      "grad_norm": 1.0683209882742306,
+      "learning_rate": 8.625132890839706e-06,
+      "loss": 1.5902,
+      "step": 932
+    },
+    {
+      "epoch": 2.2164090368608798,
+      "grad_norm": 1.3428507912542718,
+      "learning_rate": 8.606161285564826e-06,
+      "loss": 1.8124,
+      "step": 933
+    },
+    {
+      "epoch": 2.2187871581450653,
+      "grad_norm": 1.12745278115296,
+      "learning_rate": 8.587194795040286e-06,
+      "loss": 1.7277,
+      "step": 934
+    },
+    {
+      "epoch": 2.221165279429251,
+      "grad_norm": 1.214273485118672,
+      "learning_rate": 8.56823348886442e-06,
+      "loss": 1.778,
+      "step": 935
+    },
+    {
+      "epoch": 2.2235434007134365,
+      "grad_norm": 1.1850085474869128,
+      "learning_rate": 8.549277436616551e-06,
+      "loss": 1.6415,
+      "step": 936
+    },
+    {
+      "epoch": 2.225921521997622,
+      "grad_norm": 1.0025465720081261,
+      "learning_rate": 8.530326707856716e-06,
+      "loss": 1.5837,
+      "step": 937
+    },
+    {
+      "epoch": 2.228299643281807,
+      "grad_norm": 1.447315774100193,
+      "learning_rate": 8.511381372125422e-06,
+      "loss": 1.7127,
+      "step": 938
+    },
+    {
+      "epoch": 2.2306777645659928,
+      "grad_norm": 1.206599818728477,
+      "learning_rate": 8.492441498943378e-06,
+      "loss": 1.745,
+      "step": 939
+    },
+    {
+      "epoch": 2.2330558858501783,
+      "grad_norm": 1.299850116375868,
+      "learning_rate": 8.473507157811254e-06,
+      "loss": 1.645,
+      "step": 940
+    },
+    {
+      "epoch": 2.235434007134364,
+      "grad_norm": 1.0985468078995482,
+      "learning_rate": 8.454578418209421e-06,
+      "loss": 1.6661,
+      "step": 941
+    },
+    {
+      "epoch": 2.2378121284185495,
+      "grad_norm": 1.1275974183027477,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 1.7296,
+      "step": 942
+    },
+    {
+      "epoch": 2.240190249702735,
+      "grad_norm": 1.474376638809395,
+      "learning_rate": 8.416738021415069e-06,
+      "loss": 1.7114,
+      "step": 943
+    },
+    {
+      "epoch": 2.24256837098692,
+      "grad_norm": 1.243674764872066,
+      "learning_rate": 8.397826503079489e-06,
+      "loss": 1.7362,
+      "step": 944
+    },
+    {
+      "epoch": 2.2449464922711058,
+      "grad_norm": 1.217870809816204,
+      "learning_rate": 8.378920863987576e-06,
+      "loss": 1.7632,
+      "step": 945
+    },
+    {
+      "epoch": 2.2449464922711058,
+      "eval_loss": 2.4208853244781494,
+      "eval_runtime": 65.7758,
+      "eval_samples_per_second": 9.715,
+      "eval_steps_per_second": 1.216,
+      "step": 945
+    },
+    {
+      "epoch": 2.2473246135552913,
+      "grad_norm": 1.1837390067043532,
+      "learning_rate": 8.360021173514373e-06,
+      "loss": 1.802,
+      "step": 946
+    },
+    {
+      "epoch": 2.249702734839477,
+      "grad_norm": 1.1447223958071244,
+      "learning_rate": 8.341127501013096e-06,
+      "loss": 1.8955,
+      "step": 947
+    },
+    {
+      "epoch": 2.2520808561236625,
+      "grad_norm": 1.2318620845180448,
+      "learning_rate": 8.32223991581488e-06,
+      "loss": 1.7863,
+      "step": 948
+    },
+    {
+      "epoch": 2.2544589774078476,
+      "grad_norm": 1.394940898102735,
+      "learning_rate": 8.30335848722852e-06,
+      "loss": 1.7229,
+      "step": 949
+    },
+    {
+      "epoch": 2.256837098692033,
+      "grad_norm": 1.1301876418286094,
+      "learning_rate": 8.284483284540217e-06,
+      "loss": 1.7583,
+      "step": 950
+    },
+    {
+      "epoch": 2.2592152199762188,
+      "grad_norm": 1.217726889827069,
+      "learning_rate": 8.265614377013332e-06,
+      "loss": 1.5894,
+      "step": 951
+    },
+    {
+      "epoch": 2.2615933412604043,
+      "grad_norm": 1.1841467686154767,
+      "learning_rate": 8.246751833888122e-06,
+      "loss": 1.6364,
+      "step": 952
+    },
+    {
+      "epoch": 2.26397146254459,
+      "grad_norm": 1.1869481085156783,
+      "learning_rate": 8.22789572438149e-06,
+      "loss": 1.612,
+      "step": 953
+    },
+    {
+      "epoch": 2.2663495838287755,
+      "grad_norm": 1.1712814400466198,
+      "learning_rate": 8.20904611768673e-06,
+      "loss": 1.7397,
+      "step": 954
+    },
+    {
+      "epoch": 2.2687277051129606,
+      "grad_norm": 1.3442733936079978,
+      "learning_rate": 8.190203082973272e-06,
+      "loss": 1.7296,
+      "step": 955
+    },
+    {
+      "epoch": 2.271105826397146,
+      "grad_norm": 1.1056459089566495,
+      "learning_rate": 8.171366689386433e-06,
+      "loss": 1.6519,
+      "step": 956
+    },
+    {
+      "epoch": 2.2734839476813318,
+      "grad_norm": 1.1072199930457924,
+      "learning_rate": 8.152537006047161e-06,
+      "loss": 1.6024,
+      "step": 957
+    },
+    {
+      "epoch": 2.2758620689655173,
+      "grad_norm": 1.3049195548897443,
+      "learning_rate": 8.133714102051774e-06,
+      "loss": 1.825,
+      "step": 958
+    },
+    {
+      "epoch": 2.278240190249703,
+      "grad_norm": 1.1216698513231085,
+      "learning_rate": 8.114898046471721e-06,
+      "loss": 1.6569,
+      "step": 959
+    },
+    {
+      "epoch": 2.280618311533888,
+      "grad_norm": 1.3793482280459277,
+      "learning_rate": 8.096088908353316e-06,
+      "loss": 1.7476,
+      "step": 960
+    },
+    {
+      "epoch": 2.2829964328180736,
+      "grad_norm": 1.3318314124782873,
+      "learning_rate": 8.077286756717488e-06,
+      "loss": 1.7962,
+      "step": 961
+    },
+    {
+      "epoch": 2.285374554102259,
+      "grad_norm": 1.165705372513211,
+      "learning_rate": 8.058491660559536e-06,
+      "loss": 1.708,
+      "step": 962
+    },
+    {
+      "epoch": 2.287752675386445,
+      "grad_norm": 1.0416513999223338,
+      "learning_rate": 8.039703688848858e-06,
+      "loss": 1.6261,
+      "step": 963
+    },
+    {
+      "epoch": 2.2901307966706304,
+      "grad_norm": 1.2426839446823297,
+      "learning_rate": 8.020922910528717e-06,
+      "loss": 1.7174,
+      "step": 964
+    },
+    {
+      "epoch": 2.292508917954816,
+      "grad_norm": 1.2325341409873833,
+      "learning_rate": 8.002149394515973e-06,
+      "loss": 1.4361,
+      "step": 965
+    },
+    {
+      "epoch": 2.294887039239001,
+      "grad_norm": 1.055835642253928,
+      "learning_rate": 7.983383209700839e-06,
+      "loss": 1.642,
+      "step": 966
+    },
+    {
+      "epoch": 2.2972651605231866,
+      "grad_norm": 1.1036311323131265,
+      "learning_rate": 7.96462442494663e-06,
+      "loss": 1.5686,
+      "step": 967
+    },
+    {
+      "epoch": 2.299643281807372,
+      "grad_norm": 1.165719367931869,
+      "learning_rate": 7.945873109089503e-06,
+      "loss": 1.6485,
+      "step": 968
+    },
+    {
+      "epoch": 2.302021403091558,
+      "grad_norm": 1.292263443675501,
+      "learning_rate": 7.92712933093821e-06,
+      "loss": 1.4937,
+      "step": 969
+    },
+    {
+      "epoch": 2.304399524375743,
+      "grad_norm": 1.2446616344126429,
+      "learning_rate": 7.908393159273835e-06,
+      "loss": 1.61,
+      "step": 970
+    },
+    {
+      "epoch": 2.3067776456599285,
+      "grad_norm": 1.2958114102555907,
+      "learning_rate": 7.88966466284956e-06,
+      "loss": 1.7275,
+      "step": 971
+    },
+    {
+      "epoch": 2.309155766944114,
+      "grad_norm": 1.197050064334699,
+      "learning_rate": 7.870943910390392e-06,
+      "loss": 1.6454,
+      "step": 972
+    },
+    {
+      "epoch": 2.3115338882282996,
+      "grad_norm": 1.2307123050224071,
+      "learning_rate": 7.852230970592936e-06,
+      "loss": 1.6779,
+      "step": 973
+    },
+    {
+      "epoch": 2.313912009512485,
+      "grad_norm": 1.0920081628895666,
+      "learning_rate": 7.83352591212511e-06,
+      "loss": 1.6522,
+      "step": 974
+    },
+    {
+      "epoch": 2.316290130796671,
+      "grad_norm": 1.1702109811198733,
+      "learning_rate": 7.814828803625926e-06,
+      "loss": 1.5587,
+      "step": 975
+    },
+    {
+      "epoch": 2.3186682520808564,
+      "grad_norm": 1.4321131869810775,
+      "learning_rate": 7.796139713705214e-06,
+      "loss": 2.0141,
+      "step": 976
+    },
+    {
+      "epoch": 2.3210463733650415,
+      "grad_norm": 1.175162275052099,
+      "learning_rate": 7.777458710943384e-06,
+      "loss": 1.6676,
+      "step": 977
+    },
+    {
+      "epoch": 2.323424494649227,
+      "grad_norm": 1.286998127575608,
+      "learning_rate": 7.758785863891171e-06,
+      "loss": 1.7923,
+      "step": 978
+    },
+    {
+      "epoch": 2.3258026159334126,
+      "grad_norm": 1.2418872677949764,
+      "learning_rate": 7.740121241069376e-06,
+      "loss": 1.6202,
+      "step": 979
+    },
+    {
+      "epoch": 2.328180737217598,
+      "grad_norm": 1.2186583834505373,
+      "learning_rate": 7.721464910968628e-06,
+      "loss": 1.5604,
+      "step": 980
+    },
+    {
+      "epoch": 2.3305588585017833,
+      "grad_norm": 1.223129867370841,
+      "learning_rate": 7.702816942049118e-06,
+      "loss": 1.5994,
+      "step": 981
+    },
+    {
+      "epoch": 2.332936979785969,
+      "grad_norm": 1.1668829594845642,
+      "learning_rate": 7.684177402740365e-06,
+      "loss": 1.7446,
+      "step": 982
+    },
+    {
+      "epoch": 2.3353151010701545,
+      "grad_norm": 1.227131128900792,
+      "learning_rate": 7.66554636144095e-06,
+      "loss": 1.6967,
+      "step": 983
+    },
+    {
+      "epoch": 2.33769322235434,
+      "grad_norm": 1.1298937223507048,
+      "learning_rate": 7.646923886518263e-06,
+      "loss": 1.6147,
+      "step": 984
+    },
+    {
+      "epoch": 2.3400713436385256,
+      "grad_norm": 1.1172424007795727,
+      "learning_rate": 7.628310046308272e-06,
+      "loss": 1.8183,
+      "step": 985
+    },
+    {
+      "epoch": 2.342449464922711,
+      "grad_norm": 1.0995723009825376,
+      "learning_rate": 7.609704909115254e-06,
+      "loss": 1.7295,
+      "step": 986
+    },
+    {
+      "epoch": 2.344827586206897,
+      "grad_norm": 2.166347270097445,
+      "learning_rate": 7.5911085432115476e-06,
+      "loss": 1.8549,
+      "step": 987
+    },
+    {
+      "epoch": 2.347205707491082,
+      "grad_norm": 1.225917922436242,
+      "learning_rate": 7.57252101683731e-06,
+      "loss": 1.392,
+      "step": 988
+    },
+    {
+      "epoch": 2.3495838287752675,
+      "grad_norm": 1.1738557938320278,
+      "learning_rate": 7.5539423982002535e-06,
+      "loss": 1.6863,
+      "step": 989
+    },
+    {
+      "epoch": 2.351961950059453,
+      "grad_norm": 1.1432224564501046,
+      "learning_rate": 7.535372755475411e-06,
+      "loss": 1.7436,
+      "step": 990
+    },
+    {
+      "epoch": 2.3543400713436387,
+      "grad_norm": 1.1640247664089125,
+      "learning_rate": 7.516812156804874e-06,
+      "loss": 1.7646,
+      "step": 991
+    },
+    {
+      "epoch": 2.356718192627824,
+      "grad_norm": 1.2793531458628074,
+      "learning_rate": 7.4982606702975505e-06,
+      "loss": 1.6448,
+      "step": 992
+    },
+    {
+      "epoch": 2.3590963139120094,
+      "grad_norm": 1.1162056385630008,
+      "learning_rate": 7.479718364028903e-06,
+      "loss": 1.7002,
+      "step": 993
+    },
+    {
+      "epoch": 2.361474435196195,
+      "grad_norm": 1.1444550748454578,
+      "learning_rate": 7.461185306040712e-06,
+      "loss": 1.7143,
+      "step": 994
+    },
+    {
+      "epoch": 2.3638525564803805,
+      "grad_norm": 1.1918354987814666,
+      "learning_rate": 7.442661564340823e-06,
+      "loss": 1.5962,
+      "step": 995
+    },
+    {
+      "epoch": 2.366230677764566,
+      "grad_norm": 1.2320281238709774,
+      "learning_rate": 7.4241472069028915e-06,
+      "loss": 1.8181,
+      "step": 996
+    },
+    {
+      "epoch": 2.3686087990487517,
+      "grad_norm": 1.2762221347684297,
+      "learning_rate": 7.4056423016661405e-06,
+      "loss": 1.765,
+      "step": 997
+    },
+    {
+      "epoch": 2.370986920332937,
+      "grad_norm": 1.1579306898032722,
+      "learning_rate": 7.3871469165351015e-06,
+      "loss": 1.5342,
+      "step": 998
+    },
+    {
+      "epoch": 2.3733650416171224,
+      "grad_norm": 1.234792249302144,
+      "learning_rate": 7.368661119379378e-06,
+      "loss": 1.8651,
+      "step": 999
+    },
+    {
+      "epoch": 2.375743162901308,
+      "grad_norm": 1.3917341321909777,
+      "learning_rate": 7.350184978033386e-06,
+      "loss": 1.782,
+      "step": 1000
+    },
+    {
+      "epoch": 2.3781212841854935,
+      "grad_norm": 1.29943881237625,
+      "learning_rate": 7.331718560296109e-06,
+      "loss": 1.615,
+      "step": 1001
+    },
+    {
+      "epoch": 2.380499405469679,
+      "grad_norm": 1.3084311146501468,
+      "learning_rate": 7.313261933930858e-06,
+      "loss": 1.7772,
+      "step": 1002
+    },
+    {
+      "epoch": 2.382877526753864,
+      "grad_norm": 1.162552758386455,
+      "learning_rate": 7.294815166664998e-06,
+      "loss": 1.8071,
+      "step": 1003
+    },
+    {
+      "epoch": 2.38525564803805,
+      "grad_norm": 1.1614300990893485,
+      "learning_rate": 7.276378326189729e-06,
+      "loss": 1.516,
+      "step": 1004
+    },
+    {
+      "epoch": 2.3876337693222354,
+      "grad_norm": 1.154540065487417,
+      "learning_rate": 7.25795148015982e-06,
+      "loss": 1.604,
+      "step": 1005
+    },
+    {
+      "epoch": 2.390011890606421,
+      "grad_norm": 1.2082947661643852,
+      "learning_rate": 7.2395346961933645e-06,
+      "loss": 1.5811,
+      "step": 1006
+    },
+    {
+      "epoch": 2.3923900118906065,
+      "grad_norm": 1.2336019557768343,
+      "learning_rate": 7.221128041871532e-06,
+      "loss": 1.7729,
+      "step": 1007
+    },
+    {
+      "epoch": 2.394768133174792,
+      "grad_norm": 1.1152227970531388,
+      "learning_rate": 7.202731584738323e-06,
+      "loss": 1.7409,
+      "step": 1008
+    },
+    {
+      "epoch": 2.397146254458977,
+      "grad_norm": 1.3490125025605684,
+      "learning_rate": 7.184345392300318e-06,
+      "loss": 1.6814,
+      "step": 1009
+    },
+    {
+      "epoch": 2.399524375743163,
+      "grad_norm": 1.1548182248428795,
+      "learning_rate": 7.16596953202643e-06,
+      "loss": 1.6186,
+      "step": 1010
+    },
+    {
+      "epoch": 2.4019024970273484,
+      "grad_norm": 1.2852743643592273,
+      "learning_rate": 7.1476040713476605e-06,
+      "loss": 1.8935,
+      "step": 1011
+    },
+    {
+      "epoch": 2.404280618311534,
+      "grad_norm": 1.220532077802057,
+      "learning_rate": 7.129249077656844e-06,
+      "loss": 1.6207,
+      "step": 1012
+    },
+    {
+      "epoch": 2.4066587395957195,
+      "grad_norm": 1.2402232079606048,
+      "learning_rate": 7.110904618308409e-06,
+      "loss": 1.8316,
+      "step": 1013
+    },
+    {
+      "epoch": 2.4090368608799047,
+      "grad_norm": 1.1273649207025187,
+      "learning_rate": 7.092570760618132e-06,
+      "loss": 1.7531,
+      "step": 1014
+    },
+    {
+      "epoch": 2.4114149821640902,
+      "grad_norm": 1.1542485461767478,
+      "learning_rate": 7.074247571862877e-06,
+      "loss": 1.4423,
+      "step": 1015
+    },
+    {
+      "epoch": 2.413793103448276,
+      "grad_norm": 1.08202605516453,
+      "learning_rate": 7.055935119280369e-06,
+      "loss": 1.6266,
+      "step": 1016
+    },
+    {
+      "epoch": 2.4161712247324614,
+      "grad_norm": 1.1770704752495105,
+      "learning_rate": 7.0376334700689254e-06,
+      "loss": 1.6228,
+      "step": 1017
+    },
+    {
+      "epoch": 2.418549346016647,
+      "grad_norm": 1.456786984171917,
+      "learning_rate": 7.019342691387225e-06,
+      "loss": 1.7052,
+      "step": 1018
+    },
+    {
+      "epoch": 2.4209274673008325,
+      "grad_norm": 1.1838036075763612,
+      "learning_rate": 7.001062850354059e-06,
+      "loss": 1.7753,
+      "step": 1019
+    },
+    {
+      "epoch": 2.4233055885850177,
+      "grad_norm": 1.1306005750393724,
+      "learning_rate": 6.9827940140480776e-06,
+      "loss": 1.67,
+      "step": 1020
+    },
+    {
+      "epoch": 2.4256837098692032,
+      "grad_norm": 1.3893043397375318,
+      "learning_rate": 6.964536249507556e-06,
+      "loss": 1.9198,
+      "step": 1021
+    },
+    {
+      "epoch": 2.428061831153389,
+      "grad_norm": 1.1678798868137186,
+      "learning_rate": 6.946289623730131e-06,
+      "loss": 1.6164,
+      "step": 1022
+    },
+    {
+      "epoch": 2.4304399524375744,
+      "grad_norm": 1.154762934924946,
+      "learning_rate": 6.928054203672572e-06,
+      "loss": 1.7154,
+      "step": 1023
+    },
+    {
+      "epoch": 2.43281807372176,
+      "grad_norm": 1.1232082829911434,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 1.7256,
+      "step": 1024
+    },
+    {
+      "epoch": 2.435196195005945,
+      "grad_norm": 1.196133461314455,
+      "learning_rate": 6.891617248338282e-06,
+      "loss": 1.8479,
+      "step": 1025
+    },
+    {
+      "epoch": 2.4375743162901307,
+      "grad_norm": 1.2594174079784615,
+      "learning_rate": 6.873415846768503e-06,
+      "loss": 1.7605,
+      "step": 1026
+    },
+    {
+      "epoch": 2.4399524375743162,
+      "grad_norm": 1.1659462826012594,
+      "learning_rate": 6.85522591833201e-06,
+      "loss": 1.8646,
+      "step": 1027
+    },
+    {
+      "epoch": 2.442330558858502,
+      "grad_norm": 1.2914773691659898,
+      "learning_rate": 6.837047529777516e-06,
+      "loss": 1.5371,
+      "step": 1028
+    },
+    {
+      "epoch": 2.4447086801426874,
+      "grad_norm": 1.0754377635014698,
+      "learning_rate": 6.8188807478113904e-06,
+      "loss": 1.6698,
+      "step": 1029
+    },
+    {
+      "epoch": 2.447086801426873,
+      "grad_norm": 1.113563098766549,
+      "learning_rate": 6.800725639097412e-06,
+      "loss": 1.3603,
+      "step": 1030
+    },
+    {
+      "epoch": 2.449464922711058,
+      "grad_norm": 1.128889361821313,
+      "learning_rate": 6.782582270256519e-06,
+      "loss": 1.5279,
+      "step": 1031
+    },
+    {
+      "epoch": 2.4518430439952437,
+      "grad_norm": 1.329815666398893,
+      "learning_rate": 6.764450707866577e-06,
+      "loss": 1.8093,
+      "step": 1032
+    },
+    {
+      "epoch": 2.4542211652794292,
+      "grad_norm": 1.3107218996988408,
+      "learning_rate": 6.746331018462122e-06,
+      "loss": 1.6953,
+      "step": 1033
+    },
+    {
+      "epoch": 2.456599286563615,
+      "grad_norm": 1.3794075786425875,
+      "learning_rate": 6.728223268534127e-06,
+      "loss": 1.5287,
+      "step": 1034
+    },
+    {
+      "epoch": 2.4589774078478004,
+      "grad_norm": 1.2022506712726995,
+      "learning_rate": 6.710127524529746e-06,
+      "loss": 1.7323,
+      "step": 1035
+    },
+    {
+      "epoch": 2.4613555291319855,
+      "grad_norm": 1.2379907256486176,
+      "learning_rate": 6.6920438528520794e-06,
+      "loss": 1.7864,
+      "step": 1036
+    },
+    {
+      "epoch": 2.463733650416171,
+      "grad_norm": 1.2398631324380247,
+      "learning_rate": 6.673972319859928e-06,
+      "loss": 1.6605,
+      "step": 1037
+    },
+    {
+      "epoch": 2.4661117717003567,
+      "grad_norm": 1.4244397199332064,
+      "learning_rate": 6.655912991867551e-06,
+      "loss": 1.5941,
+      "step": 1038
+    },
+    {
+      "epoch": 2.4684898929845422,
+      "grad_norm": 1.0907158228131357,
+      "learning_rate": 6.6378659351444185e-06,
+      "loss": 1.6619,
+      "step": 1039
+    },
+    {
+      "epoch": 2.470868014268728,
+      "grad_norm": 1.2326470273897912,
+      "learning_rate": 6.619831215914974e-06,
+      "loss": 1.5241,
+      "step": 1040
+    },
+    {
+      "epoch": 2.4732461355529134,
+      "grad_norm": 1.1815306018044767,
+      "learning_rate": 6.601808900358382e-06,
+      "loss": 1.5294,
+      "step": 1041
+    },
+    {
+      "epoch": 2.4756242568370985,
+      "grad_norm": 1.3514710702412487,
+      "learning_rate": 6.583799054608293e-06,
+      "loss": 1.7267,
+      "step": 1042
+    },
+    {
+      "epoch": 2.478002378121284,
+      "grad_norm": 1.121258311049222,
+      "learning_rate": 6.565801744752604e-06,
+      "loss": 1.8033,
+      "step": 1043
+    },
+    {
+      "epoch": 2.4803804994054697,
+      "grad_norm": 1.0430171184122976,
+      "learning_rate": 6.547817036833208e-06,
+      "loss": 1.6306,
+      "step": 1044
+    },
+    {
+      "epoch": 2.4827586206896552,
+      "grad_norm": 1.1509719987044542,
+      "learning_rate": 6.529844996845751e-06,
+      "loss": 1.7799,
+      "step": 1045
+    },
+    {
+      "epoch": 2.485136741973841,
+      "grad_norm": 1.3436366916936602,
+      "learning_rate": 6.511885690739399e-06,
+      "loss": 1.7702,
+      "step": 1046
+    },
+    {
+      "epoch": 2.487514863258026,
+      "grad_norm": 1.2980206283561453,
+      "learning_rate": 6.4939391844165865e-06,
+      "loss": 1.6951,
+      "step": 1047
+    },
+    {
+      "epoch": 2.4898929845422115,
+      "grad_norm": 1.2496756975636276,
+      "learning_rate": 6.476005543732783e-06,
+      "loss": 1.6498,
+      "step": 1048
+    },
+    {
+      "epoch": 2.492271105826397,
+      "grad_norm": 1.5028507878209147,
+      "learning_rate": 6.4580848344962435e-06,
+      "loss": 1.6611,
+      "step": 1049
+    },
+    {
+      "epoch": 2.4946492271105827,
+      "grad_norm": 1.2800840762788763,
+      "learning_rate": 6.440177122467769e-06,
+      "loss": 1.8898,
+      "step": 1050
+    },
+    {
+      "epoch": 2.4946492271105827,
+      "eval_loss": 2.416064977645874,
+      "eval_runtime": 66.0363,
+      "eval_samples_per_second": 9.676,
+      "eval_steps_per_second": 1.211,
+      "step": 1050
+    },
+    {
+      "epoch": 2.4970273483947683,
+      "grad_norm": 1.1873534107751536,
+      "learning_rate": 6.422282473360471e-06,
+      "loss": 1.6971,
+      "step": 1051
+    },
+    {
+      "epoch": 2.499405469678954,
+      "grad_norm": 1.1655553131443337,
+      "learning_rate": 6.404400952839522e-06,
+      "loss": 1.6465,
+      "step": 1052
+    },
+    {
+      "epoch": 2.501783590963139,
+      "grad_norm": 1.3302997667501615,
+      "learning_rate": 6.386532626521924e-06,
+      "loss": 1.5254,
+      "step": 1053
+    },
+    {
+      "epoch": 2.5041617122473245,
+      "grad_norm": 1.0711049335495844,
+      "learning_rate": 6.36867755997626e-06,
+      "loss": 1.674,
+      "step": 1054
+    },
+    {
+      "epoch": 2.50653983353151,
+      "grad_norm": 1.1545823396692283,
+      "learning_rate": 6.350835818722449e-06,
+      "loss": 1.5929,
+      "step": 1055
+    },
+    {
+      "epoch": 2.5089179548156957,
+      "grad_norm": 1.329273500711881,
+      "learning_rate": 6.333007468231521e-06,
+      "loss": 1.8369,
+      "step": 1056
+    },
+    {
+      "epoch": 2.5112960760998813,
+      "grad_norm": 1.1793439619345145,
+      "learning_rate": 6.315192573925366e-06,
+      "loss": 1.8477,
+      "step": 1057
+    },
+    {
+      "epoch": 2.5136741973840664,
+      "grad_norm": 1.291888497964125,
+      "learning_rate": 6.297391201176495e-06,
+      "loss": 1.7726,
+      "step": 1058
+    },
+    {
+      "epoch": 2.516052318668252,
+      "grad_norm": 1.4228282247113229,
+      "learning_rate": 6.2796034153077976e-06,
+      "loss": 1.8691,
+      "step": 1059
+    },
+    {
+      "epoch": 2.5184304399524375,
+      "grad_norm": 1.1530624413295318,
+      "learning_rate": 6.261829281592313e-06,
+      "loss": 1.7016,
+      "step": 1060
+    },
+    {
+      "epoch": 2.520808561236623,
+      "grad_norm": 1.1299743740921777,
+      "learning_rate": 6.244068865252979e-06,
+      "loss": 1.599,
+      "step": 1061
+    },
+    {
+      "epoch": 2.5231866825208087,
+      "grad_norm": 1.626142124032786,
+      "learning_rate": 6.226322231462394e-06,
+      "loss": 2.0729,
+      "step": 1062
+    },
+    {
+      "epoch": 2.5255648038049943,
+      "grad_norm": 1.2190327254010218,
+      "learning_rate": 6.20858944534259e-06,
+      "loss": 1.6418,
+      "step": 1063
+    },
+    {
+      "epoch": 2.5279429250891794,
+      "grad_norm": 1.4103004422055179,
+      "learning_rate": 6.1908705719647735e-06,
+      "loss": 1.6974,
+      "step": 1064
+    },
+    {
+      "epoch": 2.530321046373365,
+      "grad_norm": 1.2555931729554382,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 1.7555,
+      "step": 1065
+    },
+    {
+      "epoch": 2.5326991676575505,
+      "grad_norm": 1.3278788007077313,
+      "learning_rate": 6.155474823464446e-06,
+      "loss": 1.8747,
+      "step": 1066
+    },
+    {
+      "epoch": 2.535077288941736,
+      "grad_norm": 1.0206907165277292,
+      "learning_rate": 6.137798078228139e-06,
+      "loss": 1.7062,
+      "step": 1067
+    },
+    {
+      "epoch": 2.5374554102259212,
+      "grad_norm": 1.1678494099359702,
+      "learning_rate": 6.1201355055057486e-06,
+      "loss": 1.6618,
+      "step": 1068
+    },
+    {
+      "epoch": 2.539833531510107,
+      "grad_norm": 1.276758941048691,
+      "learning_rate": 6.1024871701108345e-06,
+      "loss": 1.5699,
+      "step": 1069
+    },
+    {
+      "epoch": 2.5422116527942924,
+      "grad_norm": 1.4965690225241113,
+      "learning_rate": 6.084853136804711e-06,
+      "loss": 1.8962,
+      "step": 1070
+    },
+    {
+      "epoch": 2.544589774078478,
+      "grad_norm": 1.2781827417360803,
+      "learning_rate": 6.067233470296216e-06,
+      "loss": 1.6127,
+      "step": 1071
+    },
+    {
+      "epoch": 2.5469678953626635,
+      "grad_norm": 1.3693753990851296,
+      "learning_rate": 6.049628235241459e-06,
+      "loss": 1.7889,
+      "step": 1072
+    },
+    {
+      "epoch": 2.549346016646849,
+      "grad_norm": 1.228145174812121,
+      "learning_rate": 6.032037496243604e-06,
+      "loss": 1.5608,
+      "step": 1073
+    },
+    {
+      "epoch": 2.5517241379310347,
+      "grad_norm": 1.2063099794472876,
+      "learning_rate": 6.014461317852606e-06,
+      "loss": 1.7348,
+      "step": 1074
+    },
+    {
+      "epoch": 2.55410225921522,
+      "grad_norm": 1.217042589994472,
+      "learning_rate": 5.996899764565005e-06,
+      "loss": 1.6992,
+      "step": 1075
+    },
+    {
+      "epoch": 2.5564803804994054,
+      "grad_norm": 1.301014047396616,
+      "learning_rate": 5.979352900823663e-06,
+      "loss": 1.8162,
+      "step": 1076
+    },
+    {
+      "epoch": 2.558858501783591,
+      "grad_norm": 1.0567226366943008,
+      "learning_rate": 5.961820791017544e-06,
+      "loss": 1.6215,
+      "step": 1077
+    },
+    {
+      "epoch": 2.5612366230677766,
+      "grad_norm": 1.1919768727324513,
+      "learning_rate": 5.9443034994814655e-06,
+      "loss": 1.7082,
+      "step": 1078
+    },
+    {
+      "epoch": 2.5636147443519617,
+      "grad_norm": 1.116399660030849,
+      "learning_rate": 5.926801090495873e-06,
+      "loss": 1.7973,
+      "step": 1079
+    },
+    {
+      "epoch": 2.5659928656361473,
+      "grad_norm": 2.0791551718173236,
+      "learning_rate": 5.9093136282866014e-06,
+      "loss": 1.593,
+      "step": 1080
+    },
+    {
+      "epoch": 2.568370986920333,
+      "grad_norm": 1.102934034890405,
+      "learning_rate": 5.891841177024635e-06,
+      "loss": 1.7117,
+      "step": 1081
+    },
+    {
+      "epoch": 2.5707491082045184,
+      "grad_norm": 1.4346461019869596,
+      "learning_rate": 5.874383800825878e-06,
+      "loss": 1.7807,
+      "step": 1082
+    },
+    {
+      "epoch": 2.573127229488704,
+      "grad_norm": 1.3171902324738824,
+      "learning_rate": 5.856941563750907e-06,
+      "loss": 1.7296,
+      "step": 1083
+    },
+    {
+      "epoch": 2.5755053507728896,
+      "grad_norm": 1.3868744564812188,
+      "learning_rate": 5.839514529804757e-06,
+      "loss": 1.84,
+      "step": 1084
+    },
+    {
+      "epoch": 2.577883472057075,
+      "grad_norm": 1.2362063868655484,
+      "learning_rate": 5.822102762936667e-06,
+      "loss": 1.4915,
+      "step": 1085
+    },
+    {
+      "epoch": 2.5802615933412603,
+      "grad_norm": 1.2404289042620804,
+      "learning_rate": 5.8047063270398566e-06,
+      "loss": 1.6799,
+      "step": 1086
+    },
+    {
+      "epoch": 2.582639714625446,
+      "grad_norm": 1.1129231989370107,
+      "learning_rate": 5.78732528595128e-06,
+      "loss": 1.7227,
+      "step": 1087
+    },
+    {
+      "epoch": 2.5850178359096314,
+      "grad_norm": 1.161432165715355,
+      "learning_rate": 5.769959703451413e-06,
+      "loss": 1.7592,
+      "step": 1088
+    },
+    {
+      "epoch": 2.587395957193817,
+      "grad_norm": 1.248731815613165,
+      "learning_rate": 5.752609643263996e-06,
+      "loss": 1.5589,
+      "step": 1089
+    },
+    {
+      "epoch": 2.589774078478002,
+      "grad_norm": 1.1857883701158243,
+      "learning_rate": 5.7352751690558025e-06,
+      "loss": 1.8265,
+      "step": 1090
+    },
+    {
+      "epoch": 2.5921521997621877,
+      "grad_norm": 1.2318137626074095,
+      "learning_rate": 5.717956344436435e-06,
+      "loss": 1.8188,
+      "step": 1091
+    },
+    {
+      "epoch": 2.5945303210463733,
+      "grad_norm": 1.2118351181535407,
+      "learning_rate": 5.700653232958047e-06,
+      "loss": 1.6086,
+      "step": 1092
+    },
+    {
+      "epoch": 2.596908442330559,
+      "grad_norm": 1.3511727528093584,
+      "learning_rate": 5.683365898115145e-06,
+      "loss": 1.8693,
+      "step": 1093
+    },
+    {
+      "epoch": 2.5992865636147444,
+      "grad_norm": 1.0915962232836285,
+      "learning_rate": 5.666094403344336e-06,
+      "loss": 1.6513,
+      "step": 1094
+    },
+    {
+      "epoch": 2.60166468489893,
+      "grad_norm": 1.1047451691863088,
+      "learning_rate": 5.6488388120241e-06,
+      "loss": 1.5876,
+      "step": 1095
+    },
+    {
+      "epoch": 2.6040428061831156,
+      "grad_norm": 1.0955354456142385,
+      "learning_rate": 5.63159918747457e-06,
+      "loss": 1.7715,
+      "step": 1096
+    },
+    {
+      "epoch": 2.6064209274673007,
+      "grad_norm": 1.4037474761385684,
+      "learning_rate": 5.6143755929572704e-06,
+      "loss": 1.9165,
+      "step": 1097
+    },
+    {
+      "epoch": 2.6087990487514863,
+      "grad_norm": 1.210647433502284,
+      "learning_rate": 5.5971680916749226e-06,
+      "loss": 1.6276,
+      "step": 1098
+    },
+    {
+      "epoch": 2.611177170035672,
+      "grad_norm": 1.3342524117099221,
+      "learning_rate": 5.579976746771179e-06,
+      "loss": 1.874,
+      "step": 1099
+    },
+    {
+      "epoch": 2.6135552913198574,
+      "grad_norm": 1.219886735930388,
+      "learning_rate": 5.562801621330402e-06,
+      "loss": 1.6268,
+      "step": 1100
+    },
+    {
+      "epoch": 2.6159334126040426,
+      "grad_norm": 1.254801496767963,
+      "learning_rate": 5.545642778377457e-06,
+      "loss": 1.7502,
+      "step": 1101
+    },
+    {
+      "epoch": 2.618311533888228,
+      "grad_norm": 1.1841203502125026,
+      "learning_rate": 5.5285002808774345e-06,
+      "loss": 1.8315,
+      "step": 1102
+    },
+    {
+      "epoch": 2.6206896551724137,
+      "grad_norm": 1.4994607858303892,
+      "learning_rate": 5.511374191735467e-06,
+      "loss": 1.8898,
+      "step": 1103
+    },
+    {
+      "epoch": 2.6230677764565993,
+      "grad_norm": 1.303031931946195,
+      "learning_rate": 5.49426457379646e-06,
+      "loss": 1.5813,
+      "step": 1104
+    },
+    {
+      "epoch": 2.625445897740785,
+      "grad_norm": 1.1926132163625953,
+      "learning_rate": 5.4771714898448816e-06,
+      "loss": 1.5727,
+      "step": 1105
+    },
+    {
+      "epoch": 2.6278240190249704,
+      "grad_norm": 1.2672315312487659,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 1.6712,
+      "step": 1106
+    },
+    {
+      "epoch": 2.630202140309156,
+      "grad_norm": 1.2855515231934755,
+      "learning_rate": 5.4430351747383045e-06,
+      "loss": 1.5207,
+      "step": 1107
+    },
+    {
+      "epoch": 2.632580261593341,
+      "grad_norm": 1.262934558418249,
+      "learning_rate": 5.425992068847965e-06,
+      "loss": 1.5892,
+      "step": 1108
+    },
+    {
+      "epoch": 2.6349583828775267,
+      "grad_norm": 1.2176816137166029,
+      "learning_rate": 5.408965747473913e-06,
+      "loss": 1.7342,
+      "step": 1109
+    },
+    {
+      "epoch": 2.6373365041617123,
+      "grad_norm": 1.2492689217919521,
+      "learning_rate": 5.391956273094952e-06,
+      "loss": 1.6075,
+      "step": 1110
+    },
+    {
+      "epoch": 2.639714625445898,
+      "grad_norm": 1.0855880724356506,
+      "learning_rate": 5.374963708128076e-06,
+      "loss": 1.6012,
+      "step": 1111
+    },
+    {
+      "epoch": 2.642092746730083,
+      "grad_norm": 1.092980654609454,
+      "learning_rate": 5.357988114928221e-06,
+      "loss": 1.79,
+      "step": 1112
+    },
+    {
+      "epoch": 2.6444708680142686,
+      "grad_norm": 1.214623091364747,
+      "learning_rate": 5.341029555788051e-06,
+      "loss": 1.6981,
+      "step": 1113
+    },
+    {
+      "epoch": 2.646848989298454,
+      "grad_norm": 1.3184091992263423,
+      "learning_rate": 5.324088092937716e-06,
+      "loss": 1.6904,
+      "step": 1114
+    },
+    {
+      "epoch": 2.6492271105826397,
+      "grad_norm": 1.1971392521428008,
+      "learning_rate": 5.30716378854463e-06,
+      "loss": 1.698,
+      "step": 1115
+    },
+    {
+      "epoch": 2.6516052318668253,
+      "grad_norm": 1.0548412087042252,
+      "learning_rate": 5.2902567047132505e-06,
+      "loss": 1.7051,
+      "step": 1116
+    },
+    {
+      "epoch": 2.653983353151011,
+      "grad_norm": 1.2246037817923938,
+      "learning_rate": 5.27336690348484e-06,
+      "loss": 1.6302,
+      "step": 1117
+    },
+    {
+      "epoch": 2.6563614744351964,
+      "grad_norm": 1.1580499998747145,
+      "learning_rate": 5.2564944468372344e-06,
+      "loss": 1.4743,
+      "step": 1118
+    },
+    {
+      "epoch": 2.6587395957193816,
+      "grad_norm": 1.0651264431180674,
+      "learning_rate": 5.239639396684641e-06,
+      "loss": 1.6479,
+      "step": 1119
+    },
+    {
+      "epoch": 2.661117717003567,
+      "grad_norm": 1.243044606481579,
+      "learning_rate": 5.22280181487737e-06,
+      "loss": 1.6342,
+      "step": 1120
+    },
+    {
+      "epoch": 2.6634958382877527,
+      "grad_norm": 1.278385211653831,
+      "learning_rate": 5.205981763201653e-06,
+      "loss": 1.4981,
+      "step": 1121
+    },
+    {
+      "epoch": 2.6658739595719383,
+      "grad_norm": 1.3141585309901718,
+      "learning_rate": 5.189179303379377e-06,
+      "loss": 1.8093,
+      "step": 1122
+    },
+    {
+      "epoch": 2.6682520808561234,
+      "grad_norm": 1.132203960933878,
+      "learning_rate": 5.172394497067881e-06,
+      "loss": 1.56,
+      "step": 1123
+    },
+    {
+      "epoch": 2.670630202140309,
+      "grad_norm": 1.1928572353153866,
+      "learning_rate": 5.155627405859731e-06,
+      "loss": 1.69,
+      "step": 1124
+    },
+    {
+      "epoch": 2.6730083234244946,
+      "grad_norm": 1.1862742147972845,
+      "learning_rate": 5.138878091282472e-06,
+      "loss": 1.7242,
+      "step": 1125
+    },
+    {
+      "epoch": 2.67538644470868,
+      "grad_norm": 1.1450522807579773,
+      "learning_rate": 5.122146614798433e-06,
+      "loss": 1.5918,
+      "step": 1126
+    },
+    {
+      "epoch": 2.6777645659928657,
+      "grad_norm": 1.1890028639136556,
+      "learning_rate": 5.105433037804473e-06,
+      "loss": 1.804,
+      "step": 1127
+    },
+    {
+      "epoch": 2.6801426872770513,
+      "grad_norm": 1.228689756576672,
+      "learning_rate": 5.088737421631767e-06,
+      "loss": 1.7795,
+      "step": 1128
+    },
+    {
+      "epoch": 2.682520808561237,
+      "grad_norm": 1.1598278737067984,
+      "learning_rate": 5.072059827545598e-06,
+      "loss": 1.7347,
+      "step": 1129
+    },
+    {
+      "epoch": 2.684898929845422,
+      "grad_norm": 1.1998086711545068,
+      "learning_rate": 5.055400316745096e-06,
+      "loss": 1.8378,
+      "step": 1130
+    },
+    {
+      "epoch": 2.6872770511296076,
+      "grad_norm": 1.13047712313628,
+      "learning_rate": 5.038758950363052e-06,
+      "loss": 1.6083,
+      "step": 1131
+    },
+    {
+      "epoch": 2.689655172413793,
+      "grad_norm": 1.2433884758695508,
+      "learning_rate": 5.0221357894656605e-06,
+      "loss": 1.5978,
+      "step": 1132
+    },
+    {
+      "epoch": 2.6920332936979787,
+      "grad_norm": 1.1378408221633547,
+      "learning_rate": 5.005530895052315e-06,
+      "loss": 1.7166,
+      "step": 1133
+    },
+    {
+      "epoch": 2.694411414982164,
+      "grad_norm": 1.2328573609700277,
+      "learning_rate": 4.988944328055386e-06,
+      "loss": 1.5597,
+      "step": 1134
+    },
+    {
+      "epoch": 2.6967895362663494,
+      "grad_norm": 1.2760668630843608,
+      "learning_rate": 4.972376149339978e-06,
+      "loss": 1.8548,
+      "step": 1135
+    },
+    {
+      "epoch": 2.699167657550535,
+      "grad_norm": 1.1500316126915893,
+      "learning_rate": 4.955826419703736e-06,
+      "loss": 1.8115,
+      "step": 1136
+    },
+    {
+      "epoch": 2.7015457788347206,
+      "grad_norm": 1.1979018516552924,
+      "learning_rate": 4.939295199876588e-06,
+      "loss": 1.6296,
+      "step": 1137
+    },
+    {
+      "epoch": 2.703923900118906,
+      "grad_norm": 1.2523452493405105,
+      "learning_rate": 4.922782550520545e-06,
+      "loss": 1.5373,
+      "step": 1138
+    },
+    {
+      "epoch": 2.7063020214030917,
+      "grad_norm": 1.105439461148688,
+      "learning_rate": 4.906288532229483e-06,
+      "loss": 1.6174,
+      "step": 1139
+    },
+    {
+      "epoch": 2.7086801426872773,
+      "grad_norm": 1.110759656877746,
+      "learning_rate": 4.889813205528895e-06,
+      "loss": 1.4889,
+      "step": 1140
+    },
+    {
+      "epoch": 2.7110582639714624,
+      "grad_norm": 1.1621478050546636,
+      "learning_rate": 4.873356630875698e-06,
+      "loss": 1.7118,
+      "step": 1141
+    },
+    {
+      "epoch": 2.713436385255648,
+      "grad_norm": 1.1168812194625557,
+      "learning_rate": 4.856918868657987e-06,
+      "loss": 1.7404,
+      "step": 1142
+    },
+    {
+      "epoch": 2.7158145065398336,
+      "grad_norm": 1.2085647298687852,
+      "learning_rate": 4.840499979194827e-06,
+      "loss": 1.5431,
+      "step": 1143
+    },
+    {
+      "epoch": 2.718192627824019,
+      "grad_norm": 1.3206971686698805,
+      "learning_rate": 4.824100022736036e-06,
+      "loss": 1.646,
+      "step": 1144
+    },
+    {
+      "epoch": 2.7205707491082043,
+      "grad_norm": 1.2005611693463547,
+      "learning_rate": 4.8077190594619425e-06,
+      "loss": 1.6168,
+      "step": 1145
+    },
+    {
+      "epoch": 2.72294887039239,
+      "grad_norm": 1.3811064744767858,
+      "learning_rate": 4.791357149483195e-06,
+      "loss": 1.9062,
+      "step": 1146
+    },
+    {
+      "epoch": 2.7253269916765754,
+      "grad_norm": 1.2484162904758584,
+      "learning_rate": 4.775014352840512e-06,
+      "loss": 1.8511,
+      "step": 1147
+    },
+    {
+      "epoch": 2.727705112960761,
+      "grad_norm": 1.2749931557296785,
+      "learning_rate": 4.758690729504478e-06,
+      "loss": 1.7171,
+      "step": 1148
+    },
+    {
+      "epoch": 2.7300832342449466,
+      "grad_norm": 1.1889630755038343,
+      "learning_rate": 4.7423863393753264e-06,
+      "loss": 1.7778,
+      "step": 1149
+    },
+    {
+      "epoch": 2.732461355529132,
+      "grad_norm": 1.4063711129472924,
+      "learning_rate": 4.726101242282708e-06,
+      "loss": 1.7224,
+      "step": 1150
+    },
+    {
+      "epoch": 2.7348394768133177,
+      "grad_norm": 1.2815162496010937,
+      "learning_rate": 4.709835497985472e-06,
+      "loss": 1.663,
+      "step": 1151
+    },
+    {
+      "epoch": 2.737217598097503,
+      "grad_norm": 1.1388809085271518,
+      "learning_rate": 4.693589166171466e-06,
+      "loss": 1.7866,
+      "step": 1152
+    },
+    {
+      "epoch": 2.7395957193816884,
+      "grad_norm": 1.2040063307183857,
+      "learning_rate": 4.677362306457288e-06,
+      "loss": 1.4829,
+      "step": 1153
+    },
+    {
+      "epoch": 2.741973840665874,
+      "grad_norm": 1.2192242287644692,
+      "learning_rate": 4.661154978388094e-06,
+      "loss": 1.5552,
+      "step": 1154
+    },
+    {
+      "epoch": 2.7443519619500596,
+      "grad_norm": 1.2655433571109214,
+      "learning_rate": 4.64496724143736e-06,
+      "loss": 1.7095,
+      "step": 1155
+    },
+    {
+      "epoch": 2.7443519619500596,
+      "eval_loss": 2.399907350540161,
+      "eval_runtime": 65.5929,
+      "eval_samples_per_second": 9.742,
+      "eval_steps_per_second": 1.22,
+      "step": 1155
+    },
+    {
+      "epoch": 2.7467300832342447,
+      "grad_norm": 1.179732883618249,
+      "learning_rate": 4.628799155006669e-06,
+      "loss": 1.6908,
+      "step": 1156
+    },
+    {
+      "epoch": 2.7491082045184303,
+      "grad_norm": 1.3989245145229374,
+      "learning_rate": 4.61265077842551e-06,
+      "loss": 1.8142,
+      "step": 1157
+    },
+    {
+      "epoch": 2.751486325802616,
+      "grad_norm": 1.1191787120080152,
+      "learning_rate": 4.596522170951028e-06,
+      "loss": 1.7384,
+      "step": 1158
+    },
+    {
+      "epoch": 2.7538644470868014,
+      "grad_norm": 1.2855659443263352,
+      "learning_rate": 4.580413391767838e-06,
+      "loss": 1.9503,
+      "step": 1159
+    },
+    {
+      "epoch": 2.756242568370987,
+      "grad_norm": 1.2264542470180242,
+      "learning_rate": 4.56432449998779e-06,
+      "loss": 1.7279,
+      "step": 1160
+    },
+    {
+      "epoch": 2.7586206896551726,
+      "grad_norm": 1.2812883169114553,
+      "learning_rate": 4.548255554649748e-06,
+      "loss": 1.6427,
+      "step": 1161
+    },
+    {
+      "epoch": 2.760998810939358,
+      "grad_norm": 1.1781486757390418,
+      "learning_rate": 4.532206614719398e-06,
+      "loss": 1.5545,
+      "step": 1162
+    },
+    {
+      "epoch": 2.7633769322235433,
+      "grad_norm": 1.2277775509340916,
+      "learning_rate": 4.5161777390889985e-06,
+      "loss": 1.7287,
+      "step": 1163
+    },
+    {
+      "epoch": 2.765755053507729,
+      "grad_norm": 1.1479936742945047,
+      "learning_rate": 4.500168986577197e-06,
+      "loss": 1.6304,
+      "step": 1164
+    },
+    {
+      "epoch": 2.7681331747919145,
+      "grad_norm": 1.1624832811933103,
+      "learning_rate": 4.484180415928786e-06,
+      "loss": 1.4935,
+      "step": 1165
+    },
+    {
+      "epoch": 2.7705112960761,
+      "grad_norm": 1.3218291013979937,
+      "learning_rate": 4.468212085814501e-06,
+      "loss": 1.7302,
+      "step": 1166
+    },
+    {
+      "epoch": 2.772889417360285,
+      "grad_norm": 1.1796953319671484,
+      "learning_rate": 4.452264054830815e-06,
+      "loss": 1.7666,
+      "step": 1167
+    },
+    {
+      "epoch": 2.7752675386444707,
+      "grad_norm": 1.0832095485324484,
+      "learning_rate": 4.436336381499701e-06,
+      "loss": 1.6282,
+      "step": 1168
+    },
+    {
+      "epoch": 2.7776456599286563,
+      "grad_norm": 1.2622320191725491,
+      "learning_rate": 4.420429124268433e-06,
+      "loss": 1.78,
+      "step": 1169
+    },
+    {
+      "epoch": 2.780023781212842,
+      "grad_norm": 1.2724707119064946,
+      "learning_rate": 4.40454234150936e-06,
+      "loss": 1.6128,
+      "step": 1170
+    },
+    {
+      "epoch": 2.7824019024970275,
+      "grad_norm": 1.2162763695266292,
+      "learning_rate": 4.388676091519717e-06,
+      "loss": 1.7501,
+      "step": 1171
+    },
+    {
+      "epoch": 2.784780023781213,
+      "grad_norm": 1.0596805705972367,
+      "learning_rate": 4.372830432521377e-06,
+      "loss": 1.509,
+      "step": 1172
+    },
+    {
+      "epoch": 2.7871581450653986,
+      "grad_norm": 1.1796451830514862,
+      "learning_rate": 4.357005422660654e-06,
+      "loss": 1.6277,
+      "step": 1173
+    },
+    {
+      "epoch": 2.7895362663495837,
+      "grad_norm": 1.1869248634096565,
+      "learning_rate": 4.341201120008101e-06,
+      "loss": 1.5832,
+      "step": 1174
+    },
+    {
+      "epoch": 2.7919143876337693,
+      "grad_norm": 1.1768008704846398,
+      "learning_rate": 4.32541758255827e-06,
+      "loss": 1.5837,
+      "step": 1175
+    },
+    {
+      "epoch": 2.794292508917955,
+      "grad_norm": 1.359133421039981,
+      "learning_rate": 4.3096548682295304e-06,
+      "loss": 1.7285,
+      "step": 1176
+    },
+    {
+      "epoch": 2.7966706302021405,
+      "grad_norm": 1.1736998914426913,
+      "learning_rate": 4.293913034863827e-06,
+      "loss": 1.678,
+      "step": 1177
+    },
+    {
+      "epoch": 2.7990487514863256,
+      "grad_norm": 1.4230563916875643,
+      "learning_rate": 4.278192140226481e-06,
+      "loss": 1.8562,
+      "step": 1178
+    },
+    {
+      "epoch": 2.801426872770511,
+      "grad_norm": 1.3115643087054751,
+      "learning_rate": 4.262492242005993e-06,
+      "loss": 1.7907,
+      "step": 1179
+    },
+    {
+      "epoch": 2.8038049940546967,
+      "grad_norm": 1.2595762549552387,
+      "learning_rate": 4.2468133978137945e-06,
+      "loss": 1.7207,
+      "step": 1180
+    },
+    {
+      "epoch": 2.8061831153388823,
+      "grad_norm": 1.1249651556654061,
+      "learning_rate": 4.23115566518408e-06,
+      "loss": 1.7445,
+      "step": 1181
+    },
+    {
+      "epoch": 2.808561236623068,
+      "grad_norm": 5.900623578238023,
+      "learning_rate": 4.215519101573561e-06,
+      "loss": 1.6315,
+      "step": 1182
+    },
+    {
+      "epoch": 2.8109393579072535,
+      "grad_norm": 0.9990899222831929,
+      "learning_rate": 4.199903764361266e-06,
+      "loss": 1.7415,
+      "step": 1183
+    },
+    {
+      "epoch": 2.8133174791914386,
+      "grad_norm": 1.1614724591043748,
+      "learning_rate": 4.1843097108483465e-06,
+      "loss": 1.6034,
+      "step": 1184
+    },
+    {
+      "epoch": 2.815695600475624,
+      "grad_norm": 1.2526358529499493,
+      "learning_rate": 4.168736998257835e-06,
+      "loss": 1.618,
+      "step": 1185
+    },
+    {
+      "epoch": 2.8180737217598097,
+      "grad_norm": 1.1343545114240972,
+      "learning_rate": 4.153185683734471e-06,
+      "loss": 1.5081,
+      "step": 1186
+    },
+    {
+      "epoch": 2.8204518430439953,
+      "grad_norm": 1.236589459861005,
+      "learning_rate": 4.137655824344456e-06,
+      "loss": 1.7708,
+      "step": 1187
+    },
+    {
+      "epoch": 2.822829964328181,
+      "grad_norm": 1.1912640761030586,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 1.6344,
+      "step": 1188
+    },
+    {
+      "epoch": 2.825208085612366,
+      "grad_norm": 1.2734504646174627,
+      "learning_rate": 4.106660698835454e-06,
+      "loss": 1.7098,
+      "step": 1189
+    },
+    {
+      "epoch": 2.8275862068965516,
+      "grad_norm": 1.4358201608935164,
+      "learning_rate": 4.091195546454398e-06,
+      "loss": 1.8293,
+      "step": 1190
+    },
+    {
+      "epoch": 2.829964328180737,
+      "grad_norm": 1.2695044259379555,
+      "learning_rate": 4.07575207668213e-06,
+      "loss": 1.7176,
+      "step": 1191
+    },
+    {
+      "epoch": 2.8323424494649228,
+      "grad_norm": 1.1791999494249772,
+      "learning_rate": 4.060330346189125e-06,
+      "loss": 1.7079,
+      "step": 1192
+    },
+    {
+      "epoch": 2.8347205707491083,
+      "grad_norm": 1.0485457720571727,
+      "learning_rate": 4.044930411566072e-06,
+      "loss": 1.5055,
+      "step": 1193
+    },
+    {
+      "epoch": 2.837098692033294,
+      "grad_norm": 1.1504512570150962,
+      "learning_rate": 4.029552329323689e-06,
+      "loss": 1.7233,
+      "step": 1194
+    },
+    {
+      "epoch": 2.839476813317479,
+      "grad_norm": 1.1903698707292614,
+      "learning_rate": 4.014196155892503e-06,
+      "loss": 1.6478,
+      "step": 1195
+    },
+    {
+      "epoch": 2.8418549346016646,
+      "grad_norm": 1.1384007435698376,
+      "learning_rate": 3.998861947622635e-06,
+      "loss": 1.6235,
+      "step": 1196
+    },
+    {
+      "epoch": 2.84423305588585,
+      "grad_norm": 1.2028078323116806,
+      "learning_rate": 3.983549760783625e-06,
+      "loss": 1.6793,
+      "step": 1197
+    },
+    {
+      "epoch": 2.8466111771700358,
+      "grad_norm": 1.1401231552934252,
+      "learning_rate": 3.968259651564183e-06,
+      "loss": 1.5737,
+      "step": 1198
+    },
+    {
+      "epoch": 2.8489892984542213,
+      "grad_norm": 1.4317094122590954,
+      "learning_rate": 3.952991676072022e-06,
+      "loss": 1.7094,
+      "step": 1199
+    },
+    {
+      "epoch": 2.8513674197384065,
+      "grad_norm": 1.1867618341971813,
+      "learning_rate": 3.937745890333623e-06,
+      "loss": 1.6757,
+      "step": 1200
+    },
+    {
+      "epoch": 2.853745541022592,
+      "grad_norm": 1.24922668225119,
+      "learning_rate": 3.92252235029404e-06,
+      "loss": 1.7348,
+      "step": 1201
+    },
+    {
+      "epoch": 2.8561236623067776,
+      "grad_norm": 2.0482492936414047,
+      "learning_rate": 3.9073211118167065e-06,
+      "loss": 1.6081,
+      "step": 1202
+    },
+    {
+      "epoch": 2.858501783590963,
+      "grad_norm": 1.3060014185372235,
+      "learning_rate": 3.8921422306832046e-06,
+      "loss": 1.8234,
+      "step": 1203
+    },
+    {
+      "epoch": 2.8608799048751488,
+      "grad_norm": 1.0794119105154323,
+      "learning_rate": 3.876985762593091e-06,
+      "loss": 1.6398,
+      "step": 1204
+    },
+    {
+      "epoch": 2.8632580261593343,
+      "grad_norm": 1.0780158189794518,
+      "learning_rate": 3.861851763163665e-06,
+      "loss": 1.7754,
+      "step": 1205
+    },
+    {
+      "epoch": 2.8656361474435195,
+      "grad_norm": 1.4861159808624935,
+      "learning_rate": 3.846740287929778e-06,
+      "loss": 1.5971,
+      "step": 1206
+    },
+    {
+      "epoch": 2.868014268727705,
+      "grad_norm": 1.2055960321740489,
+      "learning_rate": 3.831651392343634e-06,
+      "loss": 1.8145,
+      "step": 1207
+    },
+    {
+      "epoch": 2.8703923900118906,
+      "grad_norm": 1.3800077509245512,
+      "learning_rate": 3.8165851317745705e-06,
+      "loss": 1.9161,
+      "step": 1208
+    },
+    {
+      "epoch": 2.872770511296076,
+      "grad_norm": 1.33689102839319,
+      "learning_rate": 3.801541561508878e-06,
+      "loss": 1.7726,
+      "step": 1209
+    },
+    {
+      "epoch": 2.8751486325802613,
+      "grad_norm": 1.2125563028361928,
+      "learning_rate": 3.7865207367495716e-06,
+      "loss": 1.82,
+      "step": 1210
+    },
+    {
+      "epoch": 2.877526753864447,
+      "grad_norm": 1.1442793595703613,
+      "learning_rate": 3.7715227126162025e-06,
+      "loss": 1.8318,
+      "step": 1211
+    },
+    {
+      "epoch": 2.8799048751486325,
+      "grad_norm": 1.1633615922891616,
+      "learning_rate": 3.756547544144664e-06,
+      "loss": 1.8416,
+      "step": 1212
+    },
+    {
+      "epoch": 2.882282996432818,
+      "grad_norm": 1.1373224437983922,
+      "learning_rate": 3.7415952862869654e-06,
+      "loss": 1.6247,
+      "step": 1213
+    },
+    {
+      "epoch": 2.8846611177170036,
+      "grad_norm": 1.137322810651517,
+      "learning_rate": 3.7266659939110585e-06,
+      "loss": 1.5931,
+      "step": 1214
+    },
+    {
+      "epoch": 2.887039239001189,
+      "grad_norm": 1.0775318054436687,
+      "learning_rate": 3.7117597218006105e-06,
+      "loss": 1.6027,
+      "step": 1215
+    },
+    {
+      "epoch": 2.8894173602853748,
+      "grad_norm": 1.3112661099152905,
+      "learning_rate": 3.69687652465482e-06,
+      "loss": 1.6443,
+      "step": 1216
+    },
+    {
+      "epoch": 2.89179548156956,
+      "grad_norm": 1.3918606030814118,
+      "learning_rate": 3.6820164570882146e-06,
+      "loss": 1.6602,
+      "step": 1217
+    },
+    {
+      "epoch": 2.8941736028537455,
+      "grad_norm": 1.3938215922035622,
+      "learning_rate": 3.667179573630437e-06,
+      "loss": 1.6631,
+      "step": 1218
+    },
+    {
+      "epoch": 2.896551724137931,
+      "grad_norm": 1.2301421785274869,
+      "learning_rate": 3.6523659287260695e-06,
+      "loss": 1.7729,
+      "step": 1219
+    },
+    {
+      "epoch": 2.8989298454221166,
+      "grad_norm": 3.1490088690528877,
+      "learning_rate": 3.6375755767344047e-06,
+      "loss": 1.614,
+      "step": 1220
+    },
+    {
+      "epoch": 2.9013079667063018,
+      "grad_norm": 1.3157024963066564,
+      "learning_rate": 3.6228085719292647e-06,
+      "loss": 1.4488,
+      "step": 1221
+    },
+    {
+      "epoch": 2.9036860879904873,
+      "grad_norm": 1.4102599859634863,
+      "learning_rate": 3.6080649684988077e-06,
+      "loss": 1.75,
+      "step": 1222
+    },
+    {
+      "epoch": 2.906064209274673,
+      "grad_norm": 1.14892092907497,
+      "learning_rate": 3.5933448205453104e-06,
+      "loss": 1.7228,
+      "step": 1223
+    },
+    {
+      "epoch": 2.9084423305588585,
+      "grad_norm": 1.136747625427715,
+      "learning_rate": 3.578648182084975e-06,
+      "loss": 1.4841,
+      "step": 1224
+    },
+    {
+      "epoch": 2.910820451843044,
+      "grad_norm": 1.138804446403443,
+      "learning_rate": 3.5639751070477478e-06,
+      "loss": 1.4907,
+      "step": 1225
+    },
+    {
+      "epoch": 2.9131985731272296,
+      "grad_norm": 1.2075949994709174,
+      "learning_rate": 3.5493256492770935e-06,
+      "loss": 1.8892,
+      "step": 1226
+    },
+    {
+      "epoch": 2.915576694411415,
+      "grad_norm": 1.3415354960606323,
+      "learning_rate": 3.5346998625298267e-06,
+      "loss": 1.8931,
+      "step": 1227
+    },
+    {
+      "epoch": 2.9179548156956003,
+      "grad_norm": 1.087555330181847,
+      "learning_rate": 3.520097800475889e-06,
+      "loss": 1.6262,
+      "step": 1228
+    },
+    {
+      "epoch": 2.920332936979786,
+      "grad_norm": 1.2882921866824912,
+      "learning_rate": 3.505519516698165e-06,
+      "loss": 1.8168,
+      "step": 1229
+    },
+    {
+      "epoch": 2.9227110582639715,
+      "grad_norm": 1.5314068926814943,
+      "learning_rate": 3.49096506469229e-06,
+      "loss": 1.8276,
+      "step": 1230
+    },
+    {
+      "epoch": 2.925089179548157,
+      "grad_norm": 1.2493103784228665,
+      "learning_rate": 3.476434497866439e-06,
+      "loss": 1.758,
+      "step": 1231
+    },
+    {
+      "epoch": 2.927467300832342,
+      "grad_norm": 1.3521688478182825,
+      "learning_rate": 3.4619278695411495e-06,
+      "loss": 1.7994,
+      "step": 1232
+    },
+    {
+      "epoch": 2.9298454221165278,
+      "grad_norm": 1.257370521239708,
+      "learning_rate": 3.447445232949107e-06,
+      "loss": 1.9069,
+      "step": 1233
+    },
+    {
+      "epoch": 2.9322235434007133,
+      "grad_norm": 1.5545256817738884,
+      "learning_rate": 3.4329866412349578e-06,
+      "loss": 1.6923,
+      "step": 1234
+    },
+    {
+      "epoch": 2.934601664684899,
+      "grad_norm": 1.357020107953056,
+      "learning_rate": 3.418552147455125e-06,
+      "loss": 1.8293,
+      "step": 1235
+    },
+    {
+      "epoch": 2.9369797859690845,
+      "grad_norm": 1.2639785875985403,
+      "learning_rate": 3.40414180457759e-06,
+      "loss": 1.6281,
+      "step": 1236
+    },
+    {
+      "epoch": 2.93935790725327,
+      "grad_norm": 1.0769659643827696,
+      "learning_rate": 3.389755665481723e-06,
+      "loss": 1.6815,
+      "step": 1237
+    },
+    {
+      "epoch": 2.9417360285374556,
+      "grad_norm": 1.1618459096532834,
+      "learning_rate": 3.3753937829580697e-06,
+      "loss": 1.6085,
+      "step": 1238
+    },
+    {
+      "epoch": 2.9441141498216408,
+      "grad_norm": 1.1040704664620735,
+      "learning_rate": 3.3610562097081646e-06,
+      "loss": 1.7131,
+      "step": 1239
+    },
+    {
+      "epoch": 2.9464922711058263,
+      "grad_norm": 1.2015097583626706,
+      "learning_rate": 3.3467429983443477e-06,
+      "loss": 1.6711,
+      "step": 1240
+    },
+    {
+      "epoch": 2.948870392390012,
+      "grad_norm": 1.2244144308823477,
+      "learning_rate": 3.33245420138955e-06,
+      "loss": 1.714,
+      "step": 1241
+    },
+    {
+      "epoch": 2.9512485136741975,
+      "grad_norm": 1.0863527927205638,
+      "learning_rate": 3.3181898712771242e-06,
+      "loss": 1.7266,
+      "step": 1242
+    },
+    {
+      "epoch": 2.9536266349583826,
+      "grad_norm": 1.1418737529822565,
+      "learning_rate": 3.3039500603506324e-06,
+      "loss": 1.7474,
+      "step": 1243
+    },
+    {
+      "epoch": 2.956004756242568,
+      "grad_norm": 1.5000461770427274,
+      "learning_rate": 3.2897348208636616e-06,
+      "loss": 1.8399,
+      "step": 1244
+    },
+    {
+      "epoch": 2.9583828775267538,
+      "grad_norm": 1.1758761049789366,
+      "learning_rate": 3.275544204979643e-06,
+      "loss": 1.6577,
+      "step": 1245
+    },
+    {
+      "epoch": 2.9607609988109393,
+      "grad_norm": 1.1225755305291578,
+      "learning_rate": 3.2613782647716374e-06,
+      "loss": 1.6723,
+      "step": 1246
+    },
+    {
+      "epoch": 2.963139120095125,
+      "grad_norm": 1.1070527347815056,
+      "learning_rate": 3.247237052222172e-06,
+      "loss": 1.4031,
+      "step": 1247
+    },
+    {
+      "epoch": 2.9655172413793105,
+      "grad_norm": 1.35321719646625,
+      "learning_rate": 3.233120619223021e-06,
+      "loss": 1.8406,
+      "step": 1248
+    },
+    {
+      "epoch": 2.967895362663496,
+      "grad_norm": 1.2364165288535383,
+      "learning_rate": 3.219029017575034e-06,
+      "loss": 1.6799,
+      "step": 1249
+    },
+    {
+      "epoch": 2.970273483947681,
+      "grad_norm": 1.2121445966430984,
+      "learning_rate": 3.2049622989879446e-06,
+      "loss": 1.8344,
+      "step": 1250
+    },
+    {
+      "epoch": 2.972651605231867,
+      "grad_norm": 1.2037195506097635,
+      "learning_rate": 3.19092051508017e-06,
+      "loss": 1.6278,
+      "step": 1251
+    },
+    {
+      "epoch": 2.9750297265160524,
+      "grad_norm": 1.2900489275647031,
+      "learning_rate": 3.1769037173786376e-06,
+      "loss": 1.6574,
+      "step": 1252
+    },
+    {
+      "epoch": 2.977407847800238,
+      "grad_norm": 1.3972721035539168,
+      "learning_rate": 3.1629119573185796e-06,
+      "loss": 1.8404,
+      "step": 1253
+    },
+    {
+      "epoch": 2.979785969084423,
+      "grad_norm": 1.332706172864179,
+      "learning_rate": 3.148945286243349e-06,
+      "loss": 1.6114,
+      "step": 1254
+    },
+    {
+      "epoch": 2.9821640903686086,
+      "grad_norm": 1.1696702073439071,
+      "learning_rate": 3.1350037554042446e-06,
+      "loss": 1.6369,
+      "step": 1255
+    },
+    {
+      "epoch": 2.984542211652794,
+      "grad_norm": 1.166223919795344,
+      "learning_rate": 3.1210874159603044e-06,
+      "loss": 1.3043,
+      "step": 1256
+    },
+    {
+      "epoch": 2.98692033293698,
+      "grad_norm": 1.2965140677555993,
+      "learning_rate": 3.1071963189781207e-06,
+      "loss": 1.6867,
+      "step": 1257
+    },
+    {
+      "epoch": 2.9892984542211654,
+      "grad_norm": 1.2624465537922454,
+      "learning_rate": 3.093330515431673e-06,
+      "loss": 1.8266,
+      "step": 1258
+    },
+    {
+      "epoch": 2.991676575505351,
+      "grad_norm": 1.4512366062742095,
+      "learning_rate": 3.079490056202107e-06,
+      "loss": 1.7582,
+      "step": 1259
+    },
+    {
+      "epoch": 2.9940546967895365,
+      "grad_norm": 1.2114616822405386,
+      "learning_rate": 3.065674992077584e-06,
+      "loss": 1.4938,
+      "step": 1260
+    },
+    {
+      "epoch": 2.9940546967895365,
+      "eval_loss": 2.375544786453247,
+      "eval_runtime": 65.6032,
+      "eval_samples_per_second": 9.74,
+      "eval_steps_per_second": 1.219,
+      "step": 1260
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1680,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 210,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.810082245674271e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}