diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4362 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9950738916256157,
+  "eval_steps": 76,
+  "global_step": 608,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003284072249589491,
+      "grad_norm": 6271.83657859153,
+      "learning_rate": 0.0,
+      "loss": 0.7244,
+      "step": 1
+    },
+    {
+      "epoch": 0.003284072249589491,
+      "eval_loss": 0.5559563636779785,
+      "eval_runtime": 253.2891,
+      "eval_samples_per_second": 8.157,
+      "eval_steps_per_second": 0.257,
+      "step": 1
+    },
+    {
+      "epoch": 0.006568144499178982,
+      "grad_norm": 5855.051575469832,
+      "learning_rate": 3.3333333333333334e-08,
+      "loss": 0.7743,
+      "step": 2
+    },
+    {
+      "epoch": 0.009852216748768473,
+      "grad_norm": 5966.931878891552,
+      "learning_rate": 6.666666666666667e-08,
+      "loss": 0.7495,
+      "step": 3
+    },
+    {
+      "epoch": 0.013136288998357963,
+      "grad_norm": 2146.4401329128254,
+      "learning_rate": 1e-07,
+      "loss": 0.7853,
+      "step": 4
+    },
+    {
+      "epoch": 0.016420361247947456,
+      "grad_norm": 1479.4537989114276,
+      "learning_rate": 1.3333333333333334e-07,
+      "loss": 0.7486,
+      "step": 5
+    },
+    {
+      "epoch": 0.019704433497536946,
+      "grad_norm": 2073.875643154414,
+      "learning_rate": 1.6666666666666665e-07,
+      "loss": 0.7395,
+      "step": 6
+    },
+    {
+      "epoch": 0.022988505747126436,
+      "grad_norm": 3306.267015593175,
+      "learning_rate": 2e-07,
+      "loss": 0.7873,
+      "step": 7
+    },
+    {
+      "epoch": 0.026272577996715927,
+      "grad_norm": 5570.427974319344,
+      "learning_rate": 2.3333333333333333e-07,
+      "loss": 0.7897,
+      "step": 8
+    },
+    {
+      "epoch": 0.029556650246305417,
+      "grad_norm": 5192.155077763508,
+      "learning_rate": 2.6666666666666667e-07,
+      "loss": 0.7411,
+      "step": 9
+    },
+    {
+      "epoch": 0.03284072249589491,
+      "grad_norm": 3385.4892242590813,
+      "learning_rate": 3e-07,
+      "loss": 0.7656,
+      "step": 10
+    },
+    {
+      "epoch": 0.0361247947454844,
+      "grad_norm": 1680.342831450108,
+      "learning_rate": 3.333333333333333e-07,
+      "loss": 0.7801,
+      "step": 11
+    },
+    {
+      "epoch": 0.03940886699507389,
+      "grad_norm": 4799.608669273176,
+      "learning_rate": 3.666666666666666e-07,
+      "loss": 0.7407,
+      "step": 12
+    },
+    {
+      "epoch": 0.042692939244663386,
+      "grad_norm": 2371.102370606964,
+      "learning_rate": 4e-07,
+      "loss": 0.7861,
+      "step": 13
+    },
+    {
+      "epoch": 0.04597701149425287,
+      "grad_norm": 22928.120442114872,
+      "learning_rate": 4.3333333333333335e-07,
+      "loss": 0.7326,
+      "step": 14
+    },
+    {
+      "epoch": 0.04926108374384237,
+      "grad_norm": 5866.1404741418755,
+      "learning_rate": 4.6666666666666666e-07,
+      "loss": 0.7236,
+      "step": 15
+    },
+    {
+      "epoch": 0.052545155993431854,
+      "grad_norm": 10335.132176960362,
+      "learning_rate": 5e-07,
+      "loss": 0.7353,
+      "step": 16
+    },
+    {
+      "epoch": 0.05582922824302135,
+      "grad_norm": 11731.708675839953,
+      "learning_rate": 5.333333333333333e-07,
+      "loss": 0.7368,
+      "step": 17
+    },
+    {
+      "epoch": 0.059113300492610835,
+      "grad_norm": 8477.833222009613,
+      "learning_rate": 5.666666666666666e-07,
+      "loss": 0.7401,
+      "step": 18
+    },
+    {
+      "epoch": 0.06239737274220033,
+      "grad_norm": 5269.214248760142,
+      "learning_rate": 6e-07,
+      "loss": 0.7378,
+      "step": 19
+    },
+    {
+      "epoch": 0.06568144499178982,
+      "grad_norm": 2906.1574452333357,
+      "learning_rate": 6.333333333333332e-07,
+      "loss": 0.7661,
+      "step": 20
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 2376.80127554879,
+      "learning_rate": 6.666666666666666e-07,
+      "loss": 0.7457,
+      "step": 21
+    },
+    {
+      "epoch": 0.0722495894909688,
+      "grad_norm": 1796.841170343826,
+      "learning_rate": 7e-07,
+      "loss": 0.7252,
+      "step": 22
+    },
+    {
+      "epoch": 0.0755336617405583,
+      "grad_norm": 4583.563298783243,
+      "learning_rate": 7.333333333333332e-07,
+      "loss": 0.7417,
+      "step": 23
+    },
+    {
+      "epoch": 0.07881773399014778,
+      "grad_norm": 5571.34039922786,
+      "learning_rate": 7.666666666666667e-07,
+      "loss": 0.7449,
+      "step": 24
+    },
+    {
+      "epoch": 0.08210180623973727,
+      "grad_norm": 1788.7468612154946,
+      "learning_rate": 8e-07,
+      "loss": 0.7515,
+      "step": 25
+    },
+    {
+      "epoch": 0.08538587848932677,
+      "grad_norm": 3448.4947901602436,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.6949,
+      "step": 26
+    },
+    {
+      "epoch": 0.08866995073891626,
+      "grad_norm": 1036.8720881203337,
+      "learning_rate": 8.666666666666667e-07,
+      "loss": 0.7494,
+      "step": 27
+    },
+    {
+      "epoch": 0.09195402298850575,
+      "grad_norm": 2185.0194951434873,
+      "learning_rate": 9e-07,
+      "loss": 0.7197,
+      "step": 28
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 2269.166582616754,
+      "learning_rate": 9.333333333333333e-07,
+      "loss": 0.7033,
+      "step": 29
+    },
+    {
+      "epoch": 0.09852216748768473,
+      "grad_norm": 11624.24252175601,
+      "learning_rate": 9.666666666666666e-07,
+      "loss": 0.6986,
+      "step": 30
+    },
+    {
+      "epoch": 0.10180623973727422,
+      "grad_norm": 4520.614291277772,
+      "learning_rate": 1e-06,
+      "loss": 0.7107,
+      "step": 31
+    },
+    {
+      "epoch": 0.10509031198686371,
+      "grad_norm": 13701.111972610783,
+      "learning_rate": 1.0333333333333333e-06,
+      "loss": 0.7159,
+      "step": 32
+    },
+    {
+      "epoch": 0.10837438423645321,
+      "grad_norm": 7449.497627677214,
+      "learning_rate": 1.0666666666666667e-06,
+      "loss": 0.7135,
+      "step": 33
+    },
+    {
+      "epoch": 0.1116584564860427,
+      "grad_norm": 2245.9408876416637,
+      "learning_rate": 1.1e-06,
+      "loss": 0.6697,
+      "step": 34
+    },
+    {
+      "epoch": 0.11494252873563218,
+      "grad_norm": 926.1867127034129,
+      "learning_rate": 1.1333333333333332e-06,
+      "loss": 0.6825,
+      "step": 35
+    },
+    {
+      "epoch": 0.11822660098522167,
+      "grad_norm": 1329.9772406233828,
+      "learning_rate": 1.1666666666666668e-06,
+      "loss": 0.6795,
+      "step": 36
+    },
+    {
+      "epoch": 0.12151067323481117,
+      "grad_norm": 1000.9622533410908,
+      "learning_rate": 1.2e-06,
+      "loss": 0.7306,
+      "step": 37
+    },
+    {
+      "epoch": 0.12479474548440066,
+      "grad_norm": 993.8780279192466,
+      "learning_rate": 1.2333333333333333e-06,
+      "loss": 0.6348,
+      "step": 38
+    },
+    {
+      "epoch": 0.12807881773399016,
+      "grad_norm": 945.1494487867217,
+      "learning_rate": 1.2666666666666665e-06,
+      "loss": 0.6355,
+      "step": 39
+    },
+    {
+      "epoch": 0.13136288998357964,
+      "grad_norm": 179.6616198157567,
+      "learning_rate": 1.3e-06,
+      "loss": 0.65,
+      "step": 40
+    },
+    {
+      "epoch": 0.13464696223316913,
+      "grad_norm": 742.7710802743131,
+      "learning_rate": 1.3333333333333332e-06,
+      "loss": 0.6482,
+      "step": 41
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 587.9958891842592,
+      "learning_rate": 1.3666666666666666e-06,
+      "loss": 0.6399,
+      "step": 42
+    },
+    {
+      "epoch": 0.1412151067323481,
+      "grad_norm": 865.3394610899527,
+      "learning_rate": 1.4e-06,
+      "loss": 0.6455,
+      "step": 43
+    },
+    {
+      "epoch": 0.1444991789819376,
+      "grad_norm": 433.76765611396485,
+      "learning_rate": 1.4333333333333333e-06,
+      "loss": 0.6863,
+      "step": 44
+    },
+    {
+      "epoch": 0.1477832512315271,
+      "grad_norm": 112.10192933210595,
+      "learning_rate": 1.4666666666666665e-06,
+      "loss": 0.645,
+      "step": 45
+    },
+    {
+      "epoch": 0.1510673234811166,
+      "grad_norm": 2442.2715722240287,
+      "learning_rate": 1.5e-06,
+      "loss": 0.6667,
+      "step": 46
+    },
+    {
+      "epoch": 0.15435139573070608,
+      "grad_norm": 319.9299141888628,
+      "learning_rate": 1.5333333333333334e-06,
+      "loss": 0.631,
+      "step": 47
+    },
+    {
+      "epoch": 0.15763546798029557,
+      "grad_norm": 158.95826958739744,
+      "learning_rate": 1.5666666666666666e-06,
+      "loss": 0.6113,
+      "step": 48
+    },
+    {
+      "epoch": 0.16091954022988506,
+      "grad_norm": 215.59873480353468,
+      "learning_rate": 1.6e-06,
+      "loss": 0.6124,
+      "step": 49
+    },
+    {
+      "epoch": 0.16420361247947454,
+      "grad_norm": 126.6284211890604,
+      "learning_rate": 1.6333333333333333e-06,
+      "loss": 0.6154,
+      "step": 50
+    },
+    {
+      "epoch": 0.16748768472906403,
+      "grad_norm": 430.61576491190806,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.6379,
+      "step": 51
+    },
+    {
+      "epoch": 0.17077175697865354,
+      "grad_norm": 338.16595207485216,
+      "learning_rate": 1.6999999999999998e-06,
+      "loss": 0.6133,
+      "step": 52
+    },
+    {
+      "epoch": 0.17405582922824303,
+      "grad_norm": 182.98276970174174,
+      "learning_rate": 1.7333333333333334e-06,
+      "loss": 0.6288,
+      "step": 53
+    },
+    {
+      "epoch": 0.17733990147783252,
+      "grad_norm": 183.8220113000725,
+      "learning_rate": 1.7666666666666666e-06,
+      "loss": 0.6234,
+      "step": 54
+    },
+    {
+      "epoch": 0.180623973727422,
+      "grad_norm": 66.92408391602912,
+      "learning_rate": 1.8e-06,
+      "loss": 0.6182,
+      "step": 55
+    },
+    {
+      "epoch": 0.1839080459770115,
+      "grad_norm": 203.51503397263764,
+      "learning_rate": 1.833333333333333e-06,
+      "loss": 0.6466,
+      "step": 56
+    },
+    {
+      "epoch": 0.18719211822660098,
+      "grad_norm": 63.90718235472869,
+      "learning_rate": 1.8666666666666667e-06,
+      "loss": 0.6321,
+      "step": 57
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 46.03571035353064,
+      "learning_rate": 1.8999999999999998e-06,
+      "loss": 0.5958,
+      "step": 58
+    },
+    {
+      "epoch": 0.19376026272577998,
+      "grad_norm": 47.046353460290206,
+      "learning_rate": 1.933333333333333e-06,
+      "loss": 0.6416,
+      "step": 59
+    },
+    {
+      "epoch": 0.19704433497536947,
+      "grad_norm": 72.74885795414816,
+      "learning_rate": 1.9666666666666663e-06,
+      "loss": 0.6279,
+      "step": 60
+    },
+    {
+      "epoch": 0.20032840722495895,
+      "grad_norm": 43.09881595994001,
+      "learning_rate": 2e-06,
+      "loss": 0.6432,
+      "step": 61
+    },
+    {
+      "epoch": 0.20361247947454844,
+      "grad_norm": 385.8401754134751,
+      "learning_rate": 1.9999835673561283e-06,
+      "loss": 0.6153,
+      "step": 62
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 55.270632684943024,
+      "learning_rate": 1.999934269964577e-06,
+      "loss": 0.606,
+      "step": 63
+    },
+    {
+      "epoch": 0.21018062397372742,
+      "grad_norm": 112.35685565413114,
+      "learning_rate": 1.9998521094455194e-06,
+      "loss": 0.6397,
+      "step": 64
+    },
+    {
+      "epoch": 0.2134646962233169,
+      "grad_norm": 145.4743296112891,
+      "learning_rate": 1.999737088499184e-06,
+      "loss": 0.6133,
+      "step": 65
+    },
+    {
+      "epoch": 0.21674876847290642,
+      "grad_norm": 92.85749788738731,
+      "learning_rate": 1.9995892109057674e-06,
+      "loss": 0.59,
+      "step": 66
+    },
+    {
+      "epoch": 0.2200328407224959,
+      "grad_norm": 119.32804840506846,
+      "learning_rate": 1.9994084815253095e-06,
+      "loss": 0.6377,
+      "step": 67
+    },
+    {
+      "epoch": 0.2233169129720854,
+      "grad_norm": 337.56982786001447,
+      "learning_rate": 1.9991949062975332e-06,
+      "loss": 0.5827,
+      "step": 68
+    },
+    {
+      "epoch": 0.22660098522167488,
+      "grad_norm": 159.57391171072265,
+      "learning_rate": 1.99894849224165e-06,
+      "loss": 0.5926,
+      "step": 69
+    },
+    {
+      "epoch": 0.22988505747126436,
+      "grad_norm": 94.69383065711324,
+      "learning_rate": 1.998669247456129e-06,
+      "loss": 0.6249,
+      "step": 70
+    },
+    {
+      "epoch": 0.23316912972085385,
+      "grad_norm": 30.865025764067703,
+      "learning_rate": 1.9983571811184294e-06,
+      "loss": 0.615,
+      "step": 71
+    },
+    {
+      "epoch": 0.23645320197044334,
+      "grad_norm": 365.12385141652493,
+      "learning_rate": 1.9980123034847023e-06,
+      "loss": 0.6231,
+      "step": 72
+    },
+    {
+      "epoch": 0.23973727422003285,
+      "grad_norm": 24.986994644338345,
+      "learning_rate": 1.99763462588945e-06,
+      "loss": 0.6005,
+      "step": 73
+    },
+    {
+      "epoch": 0.24302134646962234,
+      "grad_norm": 156.97247949274475,
+      "learning_rate": 1.997224160745155e-06,
+      "loss": 0.5986,
+      "step": 74
+    },
+    {
+      "epoch": 0.24630541871921183,
+      "grad_norm": 104.35232755227744,
+      "learning_rate": 1.9967809215418726e-06,
+      "loss": 0.616,
+      "step": 75
+    },
+    {
+      "epoch": 0.24958949096880131,
+      "grad_norm": 174.47979716876907,
+      "learning_rate": 1.996304922846787e-06,
+      "loss": 0.5997,
+      "step": 76
+    },
+    {
+      "epoch": 0.24958949096880131,
+      "eval_loss": 0.37612438201904297,
+      "eval_runtime": 252.9482,
+      "eval_samples_per_second": 8.168,
+      "eval_steps_per_second": 0.257,
+      "step": 76
+    },
+    {
+      "epoch": 0.25287356321839083,
+      "grad_norm": 633.6612254567843,
+      "learning_rate": 1.9957961803037326e-06,
+      "loss": 0.5899,
+      "step": 77
+    },
+    {
+      "epoch": 0.2561576354679803,
+      "grad_norm": 138.79302081360208,
+      "learning_rate": 1.995254710632678e-06,
+      "loss": 0.638,
+      "step": 78
+    },
+    {
+      "epoch": 0.2594417077175698,
+      "grad_norm": 17.771737220416696,
+      "learning_rate": 1.994680531629181e-06,
+      "loss": 0.6313,
+      "step": 79
+    },
+    {
+      "epoch": 0.2627257799671593,
+      "grad_norm": 123.9631655047847,
+      "learning_rate": 1.9940736621637997e-06,
+      "loss": 0.6279,
+      "step": 80
+    },
+    {
+      "epoch": 0.2660098522167488,
+      "grad_norm": 203.7102057464755,
+      "learning_rate": 1.9934341221814736e-06,
+      "loss": 0.6131,
+      "step": 81
+    },
+    {
+      "epoch": 0.26929392446633826,
+      "grad_norm": 79.30066676169302,
+      "learning_rate": 1.992761932700868e-06,
+      "loss": 0.5929,
+      "step": 82
+    },
+    {
+      "epoch": 0.27257799671592775,
+      "grad_norm": 128.16665832374028,
+      "learning_rate": 1.9920571158136835e-06,
+      "loss": 0.6162,
+      "step": 83
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 100.90994251994853,
+      "learning_rate": 1.99131969468393e-06,
+      "loss": 0.5838,
+      "step": 84
+    },
+    {
+      "epoch": 0.2791461412151067,
+      "grad_norm": 68.97227461151734,
+      "learning_rate": 1.990549693547166e-06,
+      "loss": 0.5783,
+      "step": 85
+    },
+    {
+      "epoch": 0.2824302134646962,
+      "grad_norm": 65.63550727300574,
+      "learning_rate": 1.989747137709699e-06,
+      "loss": 0.5717,
+      "step": 86
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 105.48653502246952,
+      "learning_rate": 1.988912053547758e-06,
+      "loss": 0.6336,
+      "step": 87
+    },
+    {
+      "epoch": 0.2889983579638752,
+      "grad_norm": 151.45819328204334,
+      "learning_rate": 1.988044468506625e-06,
+      "loss": 0.5966,
+      "step": 88
+    },
+    {
+      "epoch": 0.2922824302134647,
+      "grad_norm": 73.02161989799154,
+      "learning_rate": 1.9871444110997308e-06,
+      "loss": 0.6377,
+      "step": 89
+    },
+    {
+      "epoch": 0.2955665024630542,
+      "grad_norm": 55.00481114192419,
+      "learning_rate": 1.9862119109077224e-06,
+      "loss": 0.6023,
+      "step": 90
+    },
+    {
+      "epoch": 0.2988505747126437,
+      "grad_norm": 30.849915628138252,
+      "learning_rate": 1.985246998577486e-06,
+      "loss": 0.5983,
+      "step": 91
+    },
+    {
+      "epoch": 0.3021346469622332,
+      "grad_norm": 105.68162830661484,
+      "learning_rate": 1.9842497058211427e-06,
+      "loss": 0.6191,
+      "step": 92
+    },
+    {
+      "epoch": 0.3054187192118227,
+      "grad_norm": 44.85816714268683,
+      "learning_rate": 1.9832200654150074e-06,
+      "loss": 0.6019,
+      "step": 93
+    },
+    {
+      "epoch": 0.30870279146141216,
+      "grad_norm": 94.09459294470936,
+      "learning_rate": 1.982158111198507e-06,
+      "loss": 0.6224,
+      "step": 94
+    },
+    {
+      "epoch": 0.31198686371100165,
+      "grad_norm": 72.91849721086875,
+      "learning_rate": 1.9810638780730727e-06,
+      "loss": 0.6135,
+      "step": 95
+    },
+    {
+      "epoch": 0.31527093596059114,
+      "grad_norm": 155.69433676198233,
+      "learning_rate": 1.979937402000991e-06,
+      "loss": 0.5619,
+      "step": 96
+    },
+    {
+      "epoch": 0.3185550082101806,
+      "grad_norm": 91.64919893957587,
+      "learning_rate": 1.978778720004222e-06,
+      "loss": 0.5789,
+      "step": 97
+    },
+    {
+      "epoch": 0.3218390804597701,
+      "grad_norm": 58.767110092919424,
+      "learning_rate": 1.9775878701631836e-06,
+      "loss": 0.6161,
+      "step": 98
+    },
+    {
+      "epoch": 0.3251231527093596,
+      "grad_norm": 86.39938638182066,
+      "learning_rate": 1.976364891615498e-06,
+      "loss": 0.5871,
+      "step": 99
+    },
+    {
+      "epoch": 0.3284072249589491,
+      "grad_norm": 30.43501640235004,
+      "learning_rate": 1.975109824554707e-06,
+      "loss": 0.6057,
+      "step": 100
+    },
+    {
+      "epoch": 0.33169129720853857,
+      "grad_norm": 65.6953811514428,
+      "learning_rate": 1.9738227102289505e-06,
+      "loss": 0.5836,
+      "step": 101
+    },
+    {
+      "epoch": 0.33497536945812806,
+      "grad_norm": 34.69550148683274,
+      "learning_rate": 1.972503590939612e-06,
+      "loss": 0.6239,
+      "step": 102
+    },
+    {
+      "epoch": 0.33825944170771755,
+      "grad_norm": 76.34647471342404,
+      "learning_rate": 1.971152510039926e-06,
+      "loss": 0.6006,
+      "step": 103
+    },
+    {
+      "epoch": 0.3415435139573071,
+      "grad_norm": 116.3926642735392,
+      "learning_rate": 1.9697695119335546e-06,
+      "loss": 0.5907,
+      "step": 104
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 26.910647996197213,
+      "learning_rate": 1.968354642073129e-06,
+      "loss": 0.5843,
+      "step": 105
+    },
+    {
+      "epoch": 0.34811165845648606,
+      "grad_norm": 125.40744182953463,
+      "learning_rate": 1.9669079469587545e-06,
+      "loss": 0.5907,
+      "step": 106
+    },
+    {
+      "epoch": 0.35139573070607555,
+      "grad_norm": 40.15061266508225,
+      "learning_rate": 1.965429474136482e-06,
+      "loss": 0.6412,
+      "step": 107
+    },
+    {
+      "epoch": 0.35467980295566504,
+      "grad_norm": 31.1618141731111,
+      "learning_rate": 1.963919272196746e-06,
+      "loss": 0.6126,
+      "step": 108
+    },
+    {
+      "epoch": 0.3579638752052545,
+      "grad_norm": 114.23556602704848,
+      "learning_rate": 1.962377390772768e-06,
+      "loss": 0.6034,
+      "step": 109
+    },
+    {
+      "epoch": 0.361247947454844,
+      "grad_norm": 40.58756337364396,
+      "learning_rate": 1.960803880538925e-06,
+      "loss": 0.6223,
+      "step": 110
+    },
+    {
+      "epoch": 0.3645320197044335,
+      "grad_norm": 30.422688962559107,
+      "learning_rate": 1.9591987932090833e-06,
+      "loss": 0.5825,
+      "step": 111
+    },
+    {
+      "epoch": 0.367816091954023,
+      "grad_norm": 51.96081689297722,
+      "learning_rate": 1.9575621815348996e-06,
+      "loss": 0.5922,
+      "step": 112
+    },
+    {
+      "epoch": 0.37110016420361247,
+      "grad_norm": 151.3652907897673,
+      "learning_rate": 1.9558940993040883e-06,
+      "loss": 0.6014,
+      "step": 113
+    },
+    {
+      "epoch": 0.37438423645320196,
+      "grad_norm": 179.13984258713396,
+      "learning_rate": 1.9541946013386506e-06,
+      "loss": 0.5812,
+      "step": 114
+    },
+    {
+      "epoch": 0.37766830870279144,
+      "grad_norm": 209.51648981274894,
+      "learning_rate": 1.9524637434930776e-06,
+      "loss": 0.5744,
+      "step": 115
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 77.78117906962474,
+      "learning_rate": 1.950701582652509e-06,
+      "loss": 0.5924,
+      "step": 116
+    },
+    {
+      "epoch": 0.3842364532019704,
+      "grad_norm": 56.596618911823576,
+      "learning_rate": 1.9489081767308697e-06,
+      "loss": 0.5951,
+      "step": 117
+    },
+    {
+      "epoch": 0.38752052545155996,
+      "grad_norm": 25.934351726934093,
+      "learning_rate": 1.9470835846689596e-06,
+      "loss": 0.6187,
+      "step": 118
+    },
+    {
+      "epoch": 0.39080459770114945,
+      "grad_norm": 78.54544999607046,
+      "learning_rate": 1.9452278664325227e-06,
+      "loss": 0.609,
+      "step": 119
+    },
+    {
+      "epoch": 0.39408866995073893,
+      "grad_norm": 79.08588241935573,
+      "learning_rate": 1.943341083010272e-06,
+      "loss": 0.56,
+      "step": 120
+    },
+    {
+      "epoch": 0.3973727422003284,
+      "grad_norm": 72.17985855738158,
+      "learning_rate": 1.9414232964118892e-06,
+      "loss": 0.5955,
+      "step": 121
+    },
+    {
+      "epoch": 0.4006568144499179,
+      "grad_norm": 56.33038272717607,
+      "learning_rate": 1.9394745696659807e-06,
+      "loss": 0.5745,
+      "step": 122
+    },
+    {
+      "epoch": 0.4039408866995074,
+      "grad_norm": 101.16342455706896,
+      "learning_rate": 1.9374949668180134e-06,
+      "loss": 0.5891,
+      "step": 123
+    },
+    {
+      "epoch": 0.4072249589490969,
+      "grad_norm": 67.57709272161114,
+      "learning_rate": 1.935484552928204e-06,
+      "loss": 0.6101,
+      "step": 124
+    },
+    {
+      "epoch": 0.41050903119868637,
+      "grad_norm": 22.702908704179823,
+      "learning_rate": 1.9334433940693826e-06,
+      "loss": 0.5904,
+      "step": 125
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 112.85103846233335,
+      "learning_rate": 1.9313715573248235e-06,
+      "loss": 0.5888,
+      "step": 126
+    },
+    {
+      "epoch": 0.41707717569786534,
+      "grad_norm": 74.44347415115806,
+      "learning_rate": 1.929269110786037e-06,
+      "loss": 0.6,
+      "step": 127
+    },
+    {
+      "epoch": 0.42036124794745483,
+      "grad_norm": 129.01611720733342,
+      "learning_rate": 1.9271361235505337e-06,
+      "loss": 0.5721,
+      "step": 128
+    },
+    {
+      "epoch": 0.4236453201970443,
+      "grad_norm": 21.624403814788998,
+      "learning_rate": 1.9249726657195533e-06,
+      "loss": 0.5868,
+      "step": 129
+    },
+    {
+      "epoch": 0.4269293924466338,
+      "grad_norm": 43.134924014466655,
+      "learning_rate": 1.9227788083957586e-06,
+      "loss": 0.5981,
+      "step": 130
+    },
+    {
+      "epoch": 0.4302134646962233,
+      "grad_norm": 102.45985784089031,
+      "learning_rate": 1.9205546236809032e-06,
+      "loss": 0.5906,
+      "step": 131
+    },
+    {
+      "epoch": 0.43349753694581283,
+      "grad_norm": 60.949447673351266,
+      "learning_rate": 1.9183001846734574e-06,
+      "loss": 0.6352,
+      "step": 132
+    },
+    {
+      "epoch": 0.4367816091954023,
+      "grad_norm": 401.41383723285713,
+      "learning_rate": 1.9160155654662073e-06,
+      "loss": 0.5842,
+      "step": 133
+    },
+    {
+      "epoch": 0.4400656814449918,
+      "grad_norm": 35.25770567828027,
+      "learning_rate": 1.913700841143821e-06,
+      "loss": 0.5862,
+      "step": 134
+    },
+    {
+      "epoch": 0.4433497536945813,
+      "grad_norm": 914.3282302161468,
+      "learning_rate": 1.9113560877803796e-06,
+      "loss": 0.5763,
+      "step": 135
+    },
+    {
+      "epoch": 0.4466338259441708,
+      "grad_norm": 35.54605377595188,
+      "learning_rate": 1.908981382436876e-06,
+      "loss": 0.5727,
+      "step": 136
+    },
+    {
+      "epoch": 0.44991789819376027,
+      "grad_norm": 43.4767061362858,
+      "learning_rate": 1.906576803158686e-06,
+      "loss": 0.5969,
+      "step": 137
+    },
+    {
+      "epoch": 0.45320197044334976,
+      "grad_norm": 23.151864647002014,
+      "learning_rate": 1.904142428972999e-06,
+      "loss": 0.5726,
+      "step": 138
+    },
+    {
+      "epoch": 0.45648604269293924,
+      "grad_norm": 19.480214528552803,
+      "learning_rate": 1.9016783398862226e-06,
+      "loss": 0.581,
+      "step": 139
+    },
+    {
+      "epoch": 0.45977011494252873,
+      "grad_norm": 38.790487317877094,
+      "learning_rate": 1.8991846168813544e-06,
+      "loss": 0.6188,
+      "step": 140
+    },
+    {
+      "epoch": 0.4630541871921182,
+      "grad_norm": 27.00623790662471,
+      "learning_rate": 1.8966613419153178e-06,
+      "loss": 0.6242,
+      "step": 141
+    },
+    {
+      "epoch": 0.4663382594417077,
+      "grad_norm": 22.367501221325167,
+      "learning_rate": 1.8941085979162713e-06,
+      "loss": 0.6027,
+      "step": 142
+    },
+    {
+      "epoch": 0.4696223316912972,
+      "grad_norm": 447.12696125749153,
+      "learning_rate": 1.8915264687808804e-06,
+      "loss": 0.5907,
+      "step": 143
+    },
+    {
+      "epoch": 0.4729064039408867,
+      "grad_norm": 32.44003369156479,
+      "learning_rate": 1.8889150393715625e-06,
+      "loss": 0.6062,
+      "step": 144
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 144.01477406191725,
+      "learning_rate": 1.8862743955136963e-06,
+      "loss": 0.6012,
+      "step": 145
+    },
+    {
+      "epoch": 0.4794745484400657,
+      "grad_norm": 75.34671278459541,
+      "learning_rate": 1.8836046239928022e-06,
+      "loss": 0.5843,
+      "step": 146
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 46.75651484429922,
+      "learning_rate": 1.8809058125516893e-06,
+      "loss": 0.5883,
+      "step": 147
+    },
+    {
+      "epoch": 0.4860426929392447,
+      "grad_norm": 92.53132858786016,
+      "learning_rate": 1.8781780498875723e-06,
+      "loss": 0.5989,
+      "step": 148
+    },
+    {
+      "epoch": 0.48932676518883417,
+      "grad_norm": 22.671976729298738,
+      "learning_rate": 1.875421425649156e-06,
+      "loss": 0.6094,
+      "step": 149
+    },
+    {
+      "epoch": 0.49261083743842365,
+      "grad_norm": 26.928492455992494,
+      "learning_rate": 1.8726360304336893e-06,
+      "loss": 0.5949,
+      "step": 150
+    },
+    {
+      "epoch": 0.49589490968801314,
+      "grad_norm": 286.98230638362656,
+      "learning_rate": 1.8698219557839872e-06,
+      "loss": 0.595,
+      "step": 151
+    },
+    {
+      "epoch": 0.49917898193760263,
+      "grad_norm": 79.72951321359584,
+      "learning_rate": 1.8669792941854229e-06,
+      "loss": 0.6124,
+      "step": 152
+    },
+    {
+      "epoch": 0.49917898193760263,
+      "eval_loss": 0.35242295265197754,
+      "eval_runtime": 254.4417,
+      "eval_samples_per_second": 8.12,
+      "eval_steps_per_second": 0.255,
+      "step": 152
+    },
+    {
+      "epoch": 0.5024630541871922,
+      "grad_norm": 64.33414761071143,
+      "learning_rate": 1.8641081390628876e-06,
+      "loss": 0.585,
+      "step": 153
+    },
+    {
+      "epoch": 0.5057471264367817,
+      "grad_norm": 31.517353545501283,
+      "learning_rate": 1.8612085847777212e-06,
+      "loss": 0.602,
+      "step": 154
+    },
+    {
+      "epoch": 0.5090311986863711,
+      "grad_norm": 45.827071354966,
+      "learning_rate": 1.858280726624609e-06,
+      "loss": 0.5806,
+      "step": 155
+    },
+    {
+      "epoch": 0.5123152709359606,
+      "grad_norm": 18.978499335170554,
+      "learning_rate": 1.855324660828452e-06,
+      "loss": 0.6165,
+      "step": 156
+    },
+    {
+      "epoch": 0.5155993431855501,
+      "grad_norm": 67.94220614733077,
+      "learning_rate": 1.8523404845412025e-06,
+      "loss": 0.6229,
+      "step": 157
+    },
+    {
+      "epoch": 0.5188834154351396,
+      "grad_norm": 19.582077106202128,
+      "learning_rate": 1.8493282958386739e-06,
+      "loss": 0.5532,
+      "step": 158
+    },
+    {
+      "epoch": 0.5221674876847291,
+      "grad_norm": 53.75335205438214,
+      "learning_rate": 1.846288193717314e-06,
+      "loss": 0.6104,
+      "step": 159
+    },
+    {
+      "epoch": 0.5254515599343186,
+      "grad_norm": 38.261538748017415,
+      "learning_rate": 1.8432202780909538e-06,
+      "loss": 0.5848,
+      "step": 160
+    },
+    {
+      "epoch": 0.5287356321839081,
+      "grad_norm": 222.09670680924359,
+      "learning_rate": 1.8401246497875235e-06,
+      "loss": 0.5867,
+      "step": 161
+    },
+    {
+      "epoch": 0.5320197044334976,
+      "grad_norm": 571.9505490444766,
+      "learning_rate": 1.8370014105457376e-06,
+      "loss": 0.5695,
+      "step": 162
+    },
+    {
+      "epoch": 0.535303776683087,
+      "grad_norm": 48.34961117571771,
+      "learning_rate": 1.8338506630117526e-06,
+      "loss": 0.5905,
+      "step": 163
+    },
+    {
+      "epoch": 0.5385878489326765,
+      "grad_norm": 39.05700378122741,
+      "learning_rate": 1.830672510735793e-06,
+      "loss": 0.5984,
+      "step": 164
+    },
+    {
+      "epoch": 0.541871921182266,
+      "grad_norm": 58.20418532997132,
+      "learning_rate": 1.8274670581687478e-06,
+      "loss": 0.587,
+      "step": 165
+    },
+    {
+      "epoch": 0.5451559934318555,
+      "grad_norm": 245.8655111712564,
+      "learning_rate": 1.8242344106587377e-06,
+      "loss": 0.6031,
+      "step": 166
+    },
+    {
+      "epoch": 0.548440065681445,
+      "grad_norm": 105.7509385578569,
+      "learning_rate": 1.8209746744476536e-06,
+      "loss": 0.5801,
+      "step": 167
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 97.03979541618602,
+      "learning_rate": 1.8176879566676639e-06,
+      "loss": 0.6077,
+      "step": 168
+    },
+    {
+      "epoch": 0.555008210180624,
+      "grad_norm": 14.137405437866256,
+      "learning_rate": 1.8143743653376943e-06,
+      "loss": 0.592,
+      "step": 169
+    },
+    {
+      "epoch": 0.5582922824302134,
+      "grad_norm": 46.24134681947936,
+      "learning_rate": 1.811034009359877e-06,
+      "loss": 0.5912,
+      "step": 170
+    },
+    {
+      "epoch": 0.5615763546798029,
+      "grad_norm": 101.35151290055389,
+      "learning_rate": 1.8076669985159725e-06,
+      "loss": 0.6158,
+      "step": 171
+    },
+    {
+      "epoch": 0.5648604269293924,
+      "grad_norm": 140.1896691880712,
+      "learning_rate": 1.8042734434637613e-06,
+      "loss": 0.5692,
+      "step": 172
+    },
+    {
+      "epoch": 0.5681444991789819,
+      "grad_norm": 226.4192589085135,
+      "learning_rate": 1.8008534557334063e-06,
+      "loss": 0.5754,
+      "step": 173
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 45.671389908527075,
+      "learning_rate": 1.7974071477237885e-06,
+      "loss": 0.6016,
+      "step": 174
+    },
+    {
+      "epoch": 0.5747126436781609,
+      "grad_norm": 19.244548559672378,
+      "learning_rate": 1.7939346326988125e-06,
+      "loss": 0.5594,
+      "step": 175
+    },
+    {
+      "epoch": 0.5779967159277504,
+      "grad_norm": 201.20738634541007,
+      "learning_rate": 1.7904360247836834e-06,
+      "loss": 0.5781,
+      "step": 176
+    },
+    {
+      "epoch": 0.5812807881773399,
+      "grad_norm": 41.13648160887617,
+      "learning_rate": 1.7869114389611573e-06,
+      "loss": 0.5769,
+      "step": 177
+    },
+    {
+      "epoch": 0.5845648604269293,
+      "grad_norm": 287.9411519318713,
+      "learning_rate": 1.7833609910677612e-06,
+      "loss": 0.5819,
+      "step": 178
+    },
+    {
+      "epoch": 0.5878489326765188,
+      "grad_norm": 24.18855361442465,
+      "learning_rate": 1.779784797789987e-06,
+      "loss": 0.5964,
+      "step": 179
+    },
+    {
+      "epoch": 0.5911330049261084,
+      "grad_norm": 24.012668455231225,
+      "learning_rate": 1.7761829766604554e-06,
+      "loss": 0.6134,
+      "step": 180
+    },
+    {
+      "epoch": 0.5944170771756979,
+      "grad_norm": 28.51880265050756,
+      "learning_rate": 1.772555646054055e-06,
+      "loss": 0.602,
+      "step": 181
+    },
+    {
+      "epoch": 0.5977011494252874,
+      "grad_norm": 15.794558397944432,
+      "learning_rate": 1.768902925184049e-06,
+      "loss": 0.5821,
+      "step": 182
+    },
+    {
+      "epoch": 0.6009852216748769,
+      "grad_norm": 58.628563783312366,
+      "learning_rate": 1.7652249340981604e-06,
+      "loss": 0.6285,
+      "step": 183
+    },
+    {
+      "epoch": 0.6042692939244664,
+      "grad_norm": 17.619228357058972,
+      "learning_rate": 1.7615217936746242e-06,
+      "loss": 0.596,
+      "step": 184
+    },
+    {
+      "epoch": 0.6075533661740559,
+      "grad_norm": 207.93744899824176,
+      "learning_rate": 1.7577936256182167e-06,
+      "loss": 0.5917,
+      "step": 185
+    },
+    {
+      "epoch": 0.6108374384236454,
+      "grad_norm": 41.840591224822674,
+      "learning_rate": 1.754040552456253e-06,
+      "loss": 0.5866,
+      "step": 186
+    },
+    {
+      "epoch": 0.6141215106732348,
+      "grad_norm": 93.8154760971863,
+      "learning_rate": 1.7502626975345626e-06,
+      "loss": 0.6073,
+      "step": 187
+    },
+    {
+      "epoch": 0.6174055829228243,
+      "grad_norm": 17.492683199274612,
+      "learning_rate": 1.746460185013435e-06,
+      "loss": 0.5964,
+      "step": 188
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 29.276597988664253,
+      "learning_rate": 1.742633139863538e-06,
+      "loss": 0.5754,
+      "step": 189
+    },
+    {
+      "epoch": 0.6239737274220033,
+      "grad_norm": 68.18460724104617,
+      "learning_rate": 1.7387816878618117e-06,
+      "loss": 0.5803,
+      "step": 190
+    },
+    {
+      "epoch": 0.6272577996715928,
+      "grad_norm": 56.35078039350749,
+      "learning_rate": 1.7349059555873343e-06,
+      "loss": 0.5787,
+      "step": 191
+    },
+    {
+      "epoch": 0.6305418719211823,
+      "grad_norm": 43.29495360037644,
+      "learning_rate": 1.7310060704171627e-06,
+      "loss": 0.5522,
+      "step": 192
+    },
+    {
+      "epoch": 0.6338259441707718,
+      "grad_norm": 68.5551565674711,
+      "learning_rate": 1.7270821605221446e-06,
+      "loss": 0.6055,
+      "step": 193
+    },
+    {
+      "epoch": 0.6371100164203612,
+      "grad_norm": 117.39744403334397,
+      "learning_rate": 1.723134354862708e-06,
+      "loss": 0.5713,
+      "step": 194
+    },
+    {
+      "epoch": 0.6403940886699507,
+      "grad_norm": 189.04818071139826,
+      "learning_rate": 1.7191627831846222e-06,
+      "loss": 0.5748,
+      "step": 195
+    },
+    {
+      "epoch": 0.6436781609195402,
+      "grad_norm": 74.96348879368328,
+      "learning_rate": 1.7151675760147325e-06,
+      "loss": 0.5767,
+      "step": 196
+    },
+    {
+      "epoch": 0.6469622331691297,
+      "grad_norm": 19.283409944711153,
+      "learning_rate": 1.7111488646566725e-06,
+      "loss": 0.5835,
+      "step": 197
+    },
+    {
+      "epoch": 0.6502463054187192,
+      "grad_norm": 28.125800950282475,
+      "learning_rate": 1.7071067811865474e-06,
+      "loss": 0.6006,
+      "step": 198
+    },
+    {
+      "epoch": 0.6535303776683087,
+      "grad_norm": 197.02038970341124,
+      "learning_rate": 1.7030414584485934e-06,
+      "loss": 0.6071,
+      "step": 199
+    },
+    {
+      "epoch": 0.6568144499178982,
+      "grad_norm": 39.027824525327794,
+      "learning_rate": 1.6989530300508123e-06,
+      "loss": 0.5967,
+      "step": 200
+    },
+    {
+      "epoch": 0.6600985221674877,
+      "grad_norm": 31.571940924781416,
+      "learning_rate": 1.6948416303605793e-06,
+      "loss": 0.6396,
+      "step": 201
+    },
+    {
+      "epoch": 0.6633825944170771,
+      "grad_norm": 129.62214181745804,
+      "learning_rate": 1.6907073945002288e-06,
+      "loss": 0.5821,
+      "step": 202
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 162.51175499429485,
+      "learning_rate": 1.6865504583426116e-06,
+      "loss": 0.6123,
+      "step": 203
+    },
+    {
+      "epoch": 0.6699507389162561,
+      "grad_norm": 37.68050097895022,
+      "learning_rate": 1.6823709585066306e-06,
+      "loss": 0.578,
+      "step": 204
+    },
+    {
+      "epoch": 0.6732348111658456,
+      "grad_norm": 20.017550125639158,
+      "learning_rate": 1.6781690323527509e-06,
+      "loss": 0.5854,
+      "step": 205
+    },
+    {
+      "epoch": 0.6765188834154351,
+      "grad_norm": 76.61122117606573,
+      "learning_rate": 1.6739448179784844e-06,
+      "loss": 0.589,
+      "step": 206
+    },
+    {
+      "epoch": 0.6798029556650246,
+      "grad_norm": 22.411521664367136,
+      "learning_rate": 1.6696984542138519e-06,
+      "loss": 0.5844,
+      "step": 207
+    },
+    {
+      "epoch": 0.6830870279146142,
+      "grad_norm": 98.32888397272134,
+      "learning_rate": 1.6654300806168206e-06,
+      "loss": 0.5972,
+      "step": 208
+    },
+    {
+      "epoch": 0.6863711001642037,
+      "grad_norm": 25.354018912353993,
+      "learning_rate": 1.661139837468717e-06,
+      "loss": 0.5749,
+      "step": 209
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 109.57794389864921,
+      "learning_rate": 1.6568278657696162e-06,
+      "loss": 0.5318,
+      "step": 210
+    },
+    {
+      "epoch": 0.6929392446633826,
+      "grad_norm": 184.91691946093704,
+      "learning_rate": 1.6524943072337092e-06,
+      "loss": 0.5955,
+      "step": 211
+    },
+    {
+      "epoch": 0.6962233169129721,
+      "grad_norm": 51.18780040494431,
+      "learning_rate": 1.6481393042846442e-06,
+      "loss": 0.5899,
+      "step": 212
+    },
+    {
+      "epoch": 0.6995073891625616,
+      "grad_norm": 93.66347257011115,
+      "learning_rate": 1.6437630000508462e-06,
+      "loss": 0.5723,
+      "step": 213
+    },
+    {
+      "epoch": 0.7027914614121511,
+      "grad_norm": 49.27370857815919,
+      "learning_rate": 1.6393655383608132e-06,
+      "loss": 0.6148,
+      "step": 214
+    },
+    {
+      "epoch": 0.7060755336617406,
+      "grad_norm": 25.246859061737737,
+      "learning_rate": 1.6349470637383888e-06,
+      "loss": 0.589,
+      "step": 215
+    },
+    {
+      "epoch": 0.7093596059113301,
+      "grad_norm": 112.54633084879946,
+      "learning_rate": 1.630507721398013e-06,
+      "loss": 0.6262,
+      "step": 216
+    },
+    {
+      "epoch": 0.7126436781609196,
+      "grad_norm": 59.80877460933437,
+      "learning_rate": 1.6260476572399493e-06,
+      "loss": 0.573,
+      "step": 217
+    },
+    {
+      "epoch": 0.715927750410509,
+      "grad_norm": 221.05125034165957,
+      "learning_rate": 1.6215670178454892e-06,
+      "loss": 0.6132,
+      "step": 218
+    },
+    {
+      "epoch": 0.7192118226600985,
+      "grad_norm": 32.62986909195134,
+      "learning_rate": 1.6170659504721363e-06,
+      "loss": 0.5978,
+      "step": 219
+    },
+    {
+      "epoch": 0.722495894909688,
+      "grad_norm": 79.18706822806207,
+      "learning_rate": 1.6125446030487642e-06,
+      "loss": 0.6197,
+      "step": 220
+    },
+    {
+      "epoch": 0.7257799671592775,
+      "grad_norm": 646.9711401182135,
+      "learning_rate": 1.6080031241707576e-06,
+      "loss": 0.5909,
+      "step": 221
+    },
+    {
+      "epoch": 0.729064039408867,
+      "grad_norm": 227.7039164426062,
+      "learning_rate": 1.6034416630951265e-06,
+      "loss": 0.5617,
+      "step": 222
+    },
+    {
+      "epoch": 0.7323481116584565,
+      "grad_norm": 86.24663093418314,
+      "learning_rate": 1.5988603697356009e-06,
+      "loss": 0.568,
+      "step": 223
+    },
+    {
+      "epoch": 0.735632183908046,
+      "grad_norm": 28.186646002874514,
+      "learning_rate": 1.5942593946577065e-06,
+      "loss": 0.5739,
+      "step": 224
+    },
+    {
+      "epoch": 0.7389162561576355,
+      "grad_norm": 114.91672123046412,
+      "learning_rate": 1.5896388890738127e-06,
+      "loss": 0.5854,
+      "step": 225
+    },
+    {
+      "epoch": 0.7422003284072249,
+      "grad_norm": 110.20008502969989,
+      "learning_rate": 1.5849990048381648e-06,
+      "loss": 0.5792,
+      "step": 226
+    },
+    {
+      "epoch": 0.7454844006568144,
+      "grad_norm": 118.55251219758163,
+      "learning_rate": 1.5803398944418933e-06,
+      "loss": 0.5776,
+      "step": 227
+    },
+    {
+      "epoch": 0.7487684729064039,
+      "grad_norm": 55.68698315902024,
+      "learning_rate": 1.575661711008002e-06,
+      "loss": 0.5644,
+      "step": 228
+    },
+    {
+      "epoch": 0.7487684729064039,
+      "eval_loss": 0.3468731641769409,
+      "eval_runtime": 254.0407,
+      "eval_samples_per_second": 8.133,
+      "eval_steps_per_second": 0.256,
+      "step": 228
+    },
+    {
+      "epoch": 0.7520525451559934,
+      "grad_norm": 32.913448787991214,
+      "learning_rate": 1.570964608286336e-06,
+      "loss": 0.5904,
+      "step": 229
+    },
+    {
+      "epoch": 0.7553366174055829,
+      "grad_norm": 48.81085602810963,
+      "learning_rate": 1.566248740648527e-06,
+      "loss": 0.585,
+      "step": 230
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 43.34405794275439,
+      "learning_rate": 1.5615142630829228e-06,
+      "loss": 0.5503,
+      "step": 231
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 24.78222801953508,
+      "learning_rate": 1.5567613311894907e-06,
+      "loss": 0.6158,
+      "step": 232
+    },
+    {
+      "epoch": 0.7651888341543513,
+      "grad_norm": 20.01938395622359,
+      "learning_rate": 1.5519901011747043e-06,
+      "loss": 0.5622,
+      "step": 233
+    },
+    {
+      "epoch": 0.7684729064039408,
+      "grad_norm": 128.40682635925788,
+      "learning_rate": 1.5472007298464116e-06,
+      "loss": 0.5866,
+      "step": 234
+    },
+    {
+      "epoch": 0.7717569786535303,
+      "grad_norm": 18.43305197325785,
+      "learning_rate": 1.5423933746086793e-06,
+      "loss": 0.5952,
+      "step": 235
+    },
+    {
+      "epoch": 0.7750410509031199,
+      "grad_norm": 100.43618051137392,
+      "learning_rate": 1.5375681934566202e-06,
+      "loss": 0.5882,
+      "step": 236
+    },
+    {
+      "epoch": 0.7783251231527094,
+      "grad_norm": 24.521874696776806,
+      "learning_rate": 1.5327253449712018e-06,
+      "loss": 0.5784,
+      "step": 237
+    },
+    {
+      "epoch": 0.7816091954022989,
+      "grad_norm": 123.57359702902865,
+      "learning_rate": 1.5278649883140329e-06,
+      "loss": 0.6413,
+      "step": 238
+    },
+    {
+      "epoch": 0.7848932676518884,
+      "grad_norm": 106.32565785443477,
+      "learning_rate": 1.5229872832221333e-06,
+      "loss": 0.5668,
+      "step": 239
+    },
+    {
+      "epoch": 0.7881773399014779,
+      "grad_norm": 99.10065342575062,
+      "learning_rate": 1.5180923900026845e-06,
+      "loss": 0.5928,
+      "step": 240
+    },
+    {
+      "epoch": 0.7914614121510674,
+      "grad_norm": 130.80914850175915,
+      "learning_rate": 1.513180469527761e-06,
+      "loss": 0.5926,
+      "step": 241
+    },
+    {
+      "epoch": 0.7947454844006568,
+      "grad_norm": 94.9924662311697,
+      "learning_rate": 1.5082516832290421e-06,
+      "loss": 0.6016,
+      "step": 242
+    },
+    {
+      "epoch": 0.7980295566502463,
+      "grad_norm": 197.53102698941842,
+      "learning_rate": 1.5033061930925078e-06,
+      "loss": 0.5731,
+      "step": 243
+    },
+    {
+      "epoch": 0.8013136288998358,
+      "grad_norm": 38.788949526243016,
+      "learning_rate": 1.498344161653115e-06,
+      "loss": 0.5963,
+      "step": 244
+    },
+    {
+      "epoch": 0.8045977011494253,
+      "grad_norm": 38.30742214846052,
+      "learning_rate": 1.493365751989454e-06,
+      "loss": 0.5638,
+      "step": 245
+    },
+    {
+      "epoch": 0.8078817733990148,
+      "grad_norm": 648.6594536144033,
+      "learning_rate": 1.4883711277183915e-06,
+      "loss": 0.5868,
+      "step": 246
+    },
+    {
+      "epoch": 0.8111658456486043,
+      "grad_norm": 265.95515256929093,
+      "learning_rate": 1.4833604529896908e-06,
+      "loss": 0.5512,
+      "step": 247
+    },
+    {
+      "epoch": 0.8144499178981938,
+      "grad_norm": 170.80635770181408,
+      "learning_rate": 1.478333892480619e-06,
+      "loss": 0.5994,
+      "step": 248
+    },
+    {
+      "epoch": 0.8177339901477833,
+      "grad_norm": 1398.0497669156412,
+      "learning_rate": 1.4732916113905333e-06,
+      "loss": 0.5887,
+      "step": 249
+    },
+    {
+      "epoch": 0.8210180623973727,
+      "grad_norm": 251.7115266399908,
+      "learning_rate": 1.4682337754354533e-06,
+      "loss": 0.5735,
+      "step": 250
+    },
+    {
+      "epoch": 0.8243021346469622,
+      "grad_norm": 124.88918006964796,
+      "learning_rate": 1.4631605508426121e-06,
+      "loss": 0.6044,
+      "step": 251
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 608.3413931999685,
+      "learning_rate": 1.4580721043449966e-06,
+      "loss": 0.5628,
+      "step": 252
+    },
+    {
+      "epoch": 0.8308702791461412,
+      "grad_norm": 1194.4815644571033,
+      "learning_rate": 1.4529686031758642e-06,
+      "loss": 0.5933,
+      "step": 253
+    },
+    {
+      "epoch": 0.8341543513957307,
+      "grad_norm": 113.35099537580673,
+      "learning_rate": 1.4478502150632501e-06,
+      "loss": 0.5861,
+      "step": 254
+    },
+    {
+      "epoch": 0.8374384236453202,
+      "grad_norm": 153.99285185333872,
+      "learning_rate": 1.442717108224452e-06,
+      "loss": 0.5812,
+      "step": 255
+    },
+    {
+      "epoch": 0.8407224958949097,
+      "grad_norm": 130.21821279469813,
+      "learning_rate": 1.4375694513605034e-06,
+      "loss": 0.5732,
+      "step": 256
+    },
+    {
+      "epoch": 0.8440065681444991,
+      "grad_norm": 35.99304972206288,
+      "learning_rate": 1.4324074136506282e-06,
+      "loss": 0.5845,
+      "step": 257
+    },
+    {
+      "epoch": 0.8472906403940886,
+      "grad_norm": 12.583104402647741,
+      "learning_rate": 1.4272311647466808e-06,
+      "loss": 0.6195,
+      "step": 258
+    },
+    {
+      "epoch": 0.8505747126436781,
+      "grad_norm": 161.8183385483987,
+      "learning_rate": 1.4220408747675712e-06,
+      "loss": 0.6064,
+      "step": 259
+    },
+    {
+      "epoch": 0.8538587848932676,
+      "grad_norm": 25.593864430902567,
+      "learning_rate": 1.4168367142936734e-06,
+      "loss": 0.5897,
+      "step": 260
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 60.9645827697532,
+      "learning_rate": 1.411618854361218e-06,
+      "loss": 0.6029,
+      "step": 261
+    },
+    {
+      "epoch": 0.8604269293924466,
+      "grad_norm": 313.86128927339445,
+      "learning_rate": 1.4063874664566732e-06,
+      "loss": 0.5766,
+      "step": 262
+    },
+    {
+      "epoch": 0.8637110016420362,
+      "grad_norm": 376.4774663209203,
+      "learning_rate": 1.401142722511109e-06,
+      "loss": 0.5905,
+      "step": 263
+    },
+    {
+      "epoch": 0.8669950738916257,
+      "grad_norm": 23.684255757274983,
+      "learning_rate": 1.3958847948945428e-06,
+      "loss": 0.5928,
+      "step": 264
+    },
+    {
+      "epoch": 0.8702791461412152,
+      "grad_norm": 18.327052041947013,
+      "learning_rate": 1.3906138564102792e-06,
+      "loss": 0.5659,
+      "step": 265
+    },
+    {
+      "epoch": 0.8735632183908046,
+      "grad_norm": 31.74783455878064,
+      "learning_rate": 1.3853300802892282e-06,
+      "loss": 0.5656,
+      "step": 266
+    },
+    {
+      "epoch": 0.8768472906403941,
+      "grad_norm": 419.0756338024995,
+      "learning_rate": 1.3800336401842127e-06,
+      "loss": 0.5986,
+      "step": 267
+    },
+    {
+      "epoch": 0.8801313628899836,
+      "grad_norm": 77.68111946128553,
+      "learning_rate": 1.3747247101642602e-06,
+      "loss": 0.6006,
+      "step": 268
+    },
+    {
+      "epoch": 0.8834154351395731,
+      "grad_norm": 75.41201589840144,
+      "learning_rate": 1.3694034647088838e-06,
+      "loss": 0.5778,
+      "step": 269
+    },
+    {
+      "epoch": 0.8866995073891626,
+      "grad_norm": 36.241115646339644,
+      "learning_rate": 1.3640700787023462e-06,
+      "loss": 0.565,
+      "step": 270
+    },
+    {
+      "epoch": 0.8899835796387521,
+      "grad_norm": 135.20927104693348,
+      "learning_rate": 1.3587247274279139e-06,
+      "loss": 0.5763,
+      "step": 271
+    },
+    {
+      "epoch": 0.8932676518883416,
+      "grad_norm": 87.93638862635923,
+      "learning_rate": 1.3533675865620936e-06,
+      "loss": 0.5963,
+      "step": 272
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 29.613470120000667,
+      "learning_rate": 1.3479988321688618e-06,
+      "loss": 0.577,
+      "step": 273
+    },
+    {
+      "epoch": 0.8998357963875205,
+      "grad_norm": 14.319123531102576,
+      "learning_rate": 1.3426186406938766e-06,
+      "loss": 0.5961,
+      "step": 274
+    },
+    {
+      "epoch": 0.90311986863711,
+      "grad_norm": 169.5219058735149,
+      "learning_rate": 1.337227188958679e-06,
+      "loss": 0.5897,
+      "step": 275
+    },
+    {
+      "epoch": 0.9064039408866995,
+      "grad_norm": 38.00458314499751,
+      "learning_rate": 1.331824654154881e-06,
+      "loss": 0.5613,
+      "step": 276
+    },
+    {
+      "epoch": 0.909688013136289,
+      "grad_norm": 14.378667726194587,
+      "learning_rate": 1.3264112138383443e-06,
+      "loss": 0.6258,
+      "step": 277
+    },
+    {
+      "epoch": 0.9129720853858785,
+      "grad_norm": 42.72991885435684,
+      "learning_rate": 1.320987045923342e-06,
+      "loss": 0.5773,
+      "step": 278
+    },
+    {
+      "epoch": 0.916256157635468,
+      "grad_norm": 68.40318666339633,
+      "learning_rate": 1.315552328676714e-06,
+      "loss": 0.5839,
+      "step": 279
+    },
+    {
+      "epoch": 0.9195402298850575,
+      "grad_norm": 45.180537540210565,
+      "learning_rate": 1.3101072407120055e-06,
+      "loss": 0.5955,
+      "step": 280
+    },
+    {
+      "epoch": 0.922824302134647,
+      "grad_norm": 39.16288757906327,
+      "learning_rate": 1.3046519609836e-06,
+      "loss": 0.5927,
+      "step": 281
+    },
+    {
+      "epoch": 0.9261083743842364,
+      "grad_norm": 67.73988324657877,
+      "learning_rate": 1.2991866687808353e-06,
+      "loss": 0.5799,
+      "step": 282
+    },
+    {
+      "epoch": 0.9293924466338259,
+      "grad_norm": 353.64259849704393,
+      "learning_rate": 1.2937115437221117e-06,
+      "loss": 0.5764,
+      "step": 283
+    },
+    {
+      "epoch": 0.9326765188834154,
+      "grad_norm": 54.971519691287966,
+      "learning_rate": 1.2882267657489908e-06,
+      "loss": 0.5543,
+      "step": 284
+    },
+    {
+      "epoch": 0.9359605911330049,
+      "grad_norm": 22.759273671919335,
+      "learning_rate": 1.2827325151202782e-06,
+      "loss": 0.582,
+      "step": 285
+    },
+    {
+      "epoch": 0.9392446633825944,
+      "grad_norm": 14.41964639616655,
+      "learning_rate": 1.2772289724061014e-06,
+      "loss": 0.6231,
+      "step": 286
+    },
+    {
+      "epoch": 0.9425287356321839,
+      "grad_norm": 154.5418617403585,
+      "learning_rate": 1.2717163184819759e-06,
+      "loss": 0.6038,
+      "step": 287
+    },
+    {
+      "epoch": 0.9458128078817734,
+      "grad_norm": 14.212387952750024,
+      "learning_rate": 1.2661947345228593e-06,
+      "loss": 0.5684,
+      "step": 288
+    },
+    {
+      "epoch": 0.9490968801313628,
+      "grad_norm": 123.37078826032788,
+      "learning_rate": 1.2606644019971966e-06,
+      "loss": 0.5795,
+      "step": 289
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 17.029678211667036,
+      "learning_rate": 1.255125502660958e-06,
+      "loss": 0.6055,
+      "step": 290
+    },
+    {
+      "epoch": 0.9556650246305419,
+      "grad_norm": 43.985222402630065,
+      "learning_rate": 1.2495782185516637e-06,
+      "loss": 0.5785,
+      "step": 291
+    },
+    {
+      "epoch": 0.9589490968801314,
+      "grad_norm": 26.180114326110402,
+      "learning_rate": 1.2440227319824022e-06,
+      "loss": 0.5672,
+      "step": 292
+    },
+    {
+      "epoch": 0.9622331691297209,
+      "grad_norm": 58.05962982942946,
+      "learning_rate": 1.2384592255358384e-06,
+      "loss": 0.5723,
+      "step": 293
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 96.86728362154787,
+      "learning_rate": 1.232887882058212e-06,
+      "loss": 0.5667,
+      "step": 294
+    },
+    {
+      "epoch": 0.9688013136288999,
+      "grad_norm": 31.777624654282178,
+      "learning_rate": 1.2273088846533302e-06,
+      "loss": 0.6115,
+      "step": 295
+    },
+    {
+      "epoch": 0.9720853858784894,
+      "grad_norm": 83.38296129361547,
+      "learning_rate": 1.2217224166765475e-06,
+      "loss": 0.5732,
+      "step": 296
+    },
+    {
+      "epoch": 0.9753694581280788,
+      "grad_norm": 121.56881681339887,
+      "learning_rate": 1.2161286617287418e-06,
+      "loss": 0.5954,
+      "step": 297
+    },
+    {
+      "epoch": 0.9786535303776683,
+      "grad_norm": 54.8264040669832,
+      "learning_rate": 1.2105278036502787e-06,
+      "loss": 0.5931,
+      "step": 298
+    },
+    {
+      "epoch": 0.9819376026272578,
+      "grad_norm": 21.35246678957699,
+      "learning_rate": 1.2049200265149707e-06,
+      "loss": 0.5694,
+      "step": 299
+    },
+    {
+      "epoch": 0.9852216748768473,
+      "grad_norm": 1059.6690924678123,
+      "learning_rate": 1.1993055146240272e-06,
+      "loss": 0.5658,
+      "step": 300
+    },
+    {
+      "epoch": 0.9885057471264368,
+      "grad_norm": 26.188608952399726,
+      "learning_rate": 1.1936844524999966e-06,
+      "loss": 0.5654,
+      "step": 301
+    },
+    {
+      "epoch": 0.9917898193760263,
+      "grad_norm": 21.558049696210244,
+      "learning_rate": 1.1880570248807032e-06,
+      "loss": 0.5839,
+      "step": 302
+    },
+    {
+      "epoch": 0.9950738916256158,
+      "grad_norm": 17.803809249890314,
+      "learning_rate": 1.1824234167131746e-06,
+      "loss": 0.5563,
+      "step": 303
+    },
+    {
+      "epoch": 0.9983579638752053,
+      "grad_norm": 70.52662623832066,
+      "learning_rate": 1.1767838131475654e-06,
+      "loss": 0.5626,
+      "step": 304
+    },
+    {
+      "epoch": 0.9983579638752053,
+      "eval_loss": 0.3444424271583557,
+      "eval_runtime": 255.3718,
+      "eval_samples_per_second": 8.09,
+      "eval_steps_per_second": 0.255,
+      "step": 304
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 70.52662623832066,
+      "learning_rate": 1.171138399531068e-06,
+      "loss": 0.6032,
+      "step": 305
+    },
+    {
+      "epoch": 1.0032840722495895,
+      "grad_norm": 65.13191709579829,
+      "learning_rate": 1.1654873614018266e-06,
+      "loss": 0.5511,
+      "step": 306
+    },
+    {
+      "epoch": 1.006568144499179,
+      "grad_norm": 285.9719298771201,
+      "learning_rate": 1.1598308844828345e-06,
+      "loss": 0.6026,
+      "step": 307
+    },
+    {
+      "epoch": 1.0098522167487685,
+      "grad_norm": 37.39321171527541,
+      "learning_rate": 1.154169154675834e-06,
+      "loss": 0.5785,
+      "step": 308
+    },
+    {
+      "epoch": 1.013136288998358,
+      "grad_norm": 41.68071927705891,
+      "learning_rate": 1.1485023580552039e-06,
+      "loss": 0.6114,
+      "step": 309
+    },
+    {
+      "epoch": 1.0164203612479474,
+      "grad_norm": 135.40715808644197,
+      "learning_rate": 1.1428306808618454e-06,
+      "loss": 0.5782,
+      "step": 310
+    },
+    {
+      "epoch": 1.019704433497537,
+      "grad_norm": 22.790140257232764,
+      "learning_rate": 1.137154309497062e-06,
+      "loss": 0.5687,
+      "step": 311
+    },
+    {
+      "epoch": 1.0229885057471264,
+      "grad_norm": 141.91339180106723,
+      "learning_rate": 1.131473430516432e-06,
+      "loss": 0.6137,
+      "step": 312
+    },
+    {
+      "epoch": 1.026272577996716,
+      "grad_norm": 72.39990304750096,
+      "learning_rate": 1.1257882306236775e-06,
+      "loss": 0.6173,
+      "step": 313
+    },
+    {
+      "epoch": 1.0295566502463054,
+      "grad_norm": 44.689199344971264,
+      "learning_rate": 1.1200988966645284e-06,
+      "loss": 0.5684,
+      "step": 314
+    },
+    {
+      "epoch": 1.0328407224958949,
+      "grad_norm": 18.212438241962214,
+      "learning_rate": 1.1144056156205831e-06,
+      "loss": 0.5885,
+      "step": 315
+    },
+    {
+      "epoch": 1.0361247947454844,
+      "grad_norm": 32.514249930046795,
+      "learning_rate": 1.108708574603161e-06,
+      "loss": 0.6074,
+      "step": 316
+    },
+    {
+      "epoch": 1.0394088669950738,
+      "grad_norm": 17.685539914578786,
+      "learning_rate": 1.1030079608471542e-06,
+      "loss": 0.5701,
+      "step": 317
+    },
+    {
+      "epoch": 1.0426929392446633,
+      "grad_norm": 50.50561698330298,
+      "learning_rate": 1.0973039617048747e-06,
+      "loss": 0.6148,
+      "step": 318
+    },
+    {
+      "epoch": 1.0459770114942528,
+      "grad_norm": 30.494423044684428,
+      "learning_rate": 1.0915967646398948e-06,
+      "loss": 0.5631,
+      "step": 319
+    },
+    {
+      "epoch": 1.0492610837438423,
+      "grad_norm": 43.634293011722335,
+      "learning_rate": 1.0858865572208891e-06,
+      "loss": 0.5563,
+      "step": 320
+    },
+    {
+      "epoch": 1.0525451559934318,
+      "grad_norm": 34.28951266031906,
+      "learning_rate": 1.0801735271154667e-06,
+      "loss": 0.5703,
+      "step": 321
+    },
+    {
+      "epoch": 1.0558292282430213,
+      "grad_norm": 40.35540642489708,
+      "learning_rate": 1.0744578620840063e-06,
+      "loss": 0.5682,
+      "step": 322
+    },
+    {
+      "epoch": 1.0591133004926108,
+      "grad_norm": 90.22818294715287,
+      "learning_rate": 1.068739749973484e-06,
+      "loss": 0.5676,
+      "step": 323
+    },
+    {
+      "epoch": 1.0623973727422003,
+      "grad_norm": 27.34387701783233,
+      "learning_rate": 1.0630193787112993e-06,
+      "loss": 0.5751,
+      "step": 324
+    },
+    {
+      "epoch": 1.0656814449917897,
+      "grad_norm": 56.638276613527445,
+      "learning_rate": 1.0572969362990997e-06,
+      "loss": 0.6014,
+      "step": 325
+    },
+    {
+      "epoch": 1.0689655172413792,
+      "grad_norm": 15.866790094277658,
+      "learning_rate": 1.0515726108066023e-06,
+      "loss": 0.5916,
+      "step": 326
+    },
+    {
+      "epoch": 1.0722495894909687,
+      "grad_norm": 12.950547272697525,
+      "learning_rate": 1.0458465903654105e-06,
+      "loss": 0.5734,
+      "step": 327
+    },
+    {
+      "epoch": 1.0755336617405582,
+      "grad_norm": 17.279307686255628,
+      "learning_rate": 1.0401190631628347e-06,
+      "loss": 0.5899,
+      "step": 328
+    },
+    {
+      "epoch": 1.0788177339901477,
+      "grad_norm": 26.36195986983623,
+      "learning_rate": 1.0343902174357038e-06,
+      "loss": 0.5936,
+      "step": 329
+    },
+    {
+      "epoch": 1.0821018062397372,
+      "grad_norm": 53.631919940788016,
+      "learning_rate": 1.0286602414641815e-06,
+      "loss": 0.6059,
+      "step": 330
+    },
+    {
+      "epoch": 1.0853858784893267,
+      "grad_norm": 37.558445588790654,
+      "learning_rate": 1.0229293235655768e-06,
+      "loss": 0.5573,
+      "step": 331
+    },
+    {
+      "epoch": 1.0886699507389164,
+      "grad_norm": 15.958318674221253,
+      "learning_rate": 1.017197652088155e-06,
+      "loss": 0.6102,
+      "step": 332
+    },
+    {
+      "epoch": 1.0919540229885056,
+      "grad_norm": 40.26474283910818,
+      "learning_rate": 1.0114654154049489e-06,
+      "loss": 0.5895,
+      "step": 333
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 146.08840587905024,
+      "learning_rate": 1.0057328019075668e-06,
+      "loss": 0.5726,
+      "step": 334
+    },
+    {
+      "epoch": 1.0985221674876848,
+      "grad_norm": 202.1576115663822,
+      "learning_rate": 1e-06,
+      "loss": 0.577,
+      "step": 335
+    },
+    {
+      "epoch": 1.1018062397372743,
+      "grad_norm": 14.523463101670913,
+      "learning_rate": 9.942671980924335e-07,
+      "loss": 0.5872,
+      "step": 336
+    },
+    {
+      "epoch": 1.1050903119868638,
+      "grad_norm": 32.09224127026215,
+      "learning_rate": 9.885345845950508e-07,
+      "loss": 0.6071,
+      "step": 337
+    },
+    {
+      "epoch": 1.1083743842364533,
+      "grad_norm": 13.896096566133576,
+      "learning_rate": 9.828023479118448e-07,
+      "loss": 0.5989,
+      "step": 338
+    },
+    {
+      "epoch": 1.1116584564860428,
+      "grad_norm": 58.06165751181001,
+      "learning_rate": 9.770706764344234e-07,
+      "loss": 0.5656,
+      "step": 339
+    },
+    {
+      "epoch": 1.1149425287356323,
+      "grad_norm": 20.036356068409564,
+      "learning_rate": 9.713397585358188e-07,
+      "loss": 0.5858,
+      "step": 340
+    },
+    {
+      "epoch": 1.1182266009852218,
+      "grad_norm": 10.100931882003527,
+      "learning_rate": 9.65609782564296e-07,
+      "loss": 0.5855,
+      "step": 341
+    },
+    {
+      "epoch": 1.1215106732348112,
+      "grad_norm": 13.587121183019379,
+      "learning_rate": 9.598809368371654e-07,
+      "loss": 0.6379,
+      "step": 342
+    },
+    {
+      "epoch": 1.1247947454844007,
+      "grad_norm": 126.86325897606325,
+      "learning_rate": 9.541534096345897e-07,
+      "loss": 0.5546,
+      "step": 343
+    },
+    {
+      "epoch": 1.1280788177339902,
+      "grad_norm": 169.89985357006515,
+      "learning_rate": 9.484273891933981e-07,
+      "loss": 0.5563,
+      "step": 344
+    },
+    {
+      "epoch": 1.1313628899835797,
+      "grad_norm": 49.08212390882699,
+      "learning_rate": 9.427030637009002e-07,
+      "loss": 0.5699,
+      "step": 345
+    },
+    {
+      "epoch": 1.1346469622331692,
+      "grad_norm": 40.18400735284968,
+      "learning_rate": 9.369806212887007e-07,
+      "loss": 0.5715,
+      "step": 346
+    },
+    {
+      "epoch": 1.1379310344827587,
+      "grad_norm": 18.860982747952214,
+      "learning_rate": 9.312602500265159e-07,
+      "loss": 0.5648,
+      "step": 347
+    },
+    {
+      "epoch": 1.1412151067323482,
+      "grad_norm": 10.902272643558419,
+      "learning_rate": 9.255421379159933e-07,
+      "loss": 0.5735,
+      "step": 348
+    },
+    {
+      "epoch": 1.1444991789819376,
+      "grad_norm": 66.12684936925638,
+      "learning_rate": 9.198264728845331e-07,
+      "loss": 0.6154,
+      "step": 349
+    },
+    {
+      "epoch": 1.1477832512315271,
+      "grad_norm": 28.02996822136127,
+      "learning_rate": 9.141134427791109e-07,
+      "loss": 0.5778,
+      "step": 350
+    },
+    {
+      "epoch": 1.1510673234811166,
+      "grad_norm": 30.51347632697842,
+      "learning_rate": 9.084032353601052e-07,
+      "loss": 0.6041,
+      "step": 351
+    },
+    {
+      "epoch": 1.154351395730706,
+      "grad_norm": 49.552466990210895,
+      "learning_rate": 9.026960382951252e-07,
+      "loss": 0.5671,
+      "step": 352
+    },
+    {
+      "epoch": 1.1576354679802956,
+      "grad_norm": 44.54866711025706,
+      "learning_rate": 8.969920391528457e-07,
+      "loss": 0.5506,
+      "step": 353
+    },
+    {
+      "epoch": 1.160919540229885,
+      "grad_norm": 52.856706858816,
+      "learning_rate": 8.912914253968391e-07,
+      "loss": 0.554,
+      "step": 354
+    },
+    {
+      "epoch": 1.1642036124794746,
+      "grad_norm": 47.388855420890756,
+      "learning_rate": 8.85594384379417e-07,
+      "loss": 0.5551,
+      "step": 355
+    },
+    {
+      "epoch": 1.167487684729064,
+      "grad_norm": 30.628301446970205,
+      "learning_rate": 8.799011033354715e-07,
+      "loss": 0.5831,
+      "step": 356
+    },
+    {
+      "epoch": 1.1707717569786535,
+      "grad_norm": 22.952426384099383,
+      "learning_rate": 8.742117693763227e-07,
+      "loss": 0.5597,
+      "step": 357
+    },
+    {
+      "epoch": 1.174055829228243,
+      "grad_norm": 17.939719949680224,
+      "learning_rate": 8.685265694835681e-07,
+      "loss": 0.577,
+      "step": 358
+    },
+    {
+      "epoch": 1.1773399014778325,
+      "grad_norm": 27.110777787948248,
+      "learning_rate": 8.628456905029382e-07,
+      "loss": 0.5726,
+      "step": 359
+    },
+    {
+      "epoch": 1.180623973727422,
+      "grad_norm": 13.896701402873619,
+      "learning_rate": 8.571693191381544e-07,
+      "loss": 0.5623,
+      "step": 360
+    },
+    {
+      "epoch": 1.1839080459770115,
+      "grad_norm": 15.205774656115123,
+      "learning_rate": 8.514976419447963e-07,
+      "loss": 0.5943,
+      "step": 361
+    },
+    {
+      "epoch": 1.187192118226601,
+      "grad_norm": 45.58828461857186,
+      "learning_rate": 8.458308453241663e-07,
+      "loss": 0.5817,
+      "step": 362
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 116.8441584869502,
+      "learning_rate": 8.401691155171652e-07,
+      "loss": 0.5444,
+      "step": 363
+    },
+    {
+      "epoch": 1.19376026272578,
+      "grad_norm": 170.81143104953796,
+      "learning_rate": 8.345126385981735e-07,
+      "loss": 0.5904,
+      "step": 364
+    },
+    {
+      "epoch": 1.1970443349753694,
+      "grad_norm": 41.71381416465335,
+      "learning_rate": 8.288616004689319e-07,
+      "loss": 0.5784,
+      "step": 365
+    },
+    {
+      "epoch": 1.200328407224959,
+      "grad_norm": 115.88107721819807,
+      "learning_rate": 8.23216186852435e-07,
+      "loss": 0.594,
+      "step": 366
+    },
+    {
+      "epoch": 1.2036124794745484,
+      "grad_norm": 17.490648876213996,
+      "learning_rate": 8.175765832868251e-07,
+      "loss": 0.576,
+      "step": 367
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 39.174813390569106,
+      "learning_rate": 8.11942975119297e-07,
+      "loss": 0.5627,
+      "step": 368
+    },
+    {
+      "epoch": 1.2101806239737274,
+      "grad_norm": 668.8532117187957,
+      "learning_rate": 8.063155475000035e-07,
+      "loss": 0.5987,
+      "step": 369
+    },
+    {
+      "epoch": 1.2134646962233169,
+      "grad_norm": 29.45366016595164,
+      "learning_rate": 8.006944853759732e-07,
+      "loss": 0.5788,
+      "step": 370
+    },
+    {
+      "epoch": 1.2167487684729064,
+      "grad_norm": 19.85954794293891,
+      "learning_rate": 7.950799734850291e-07,
+      "loss": 0.5552,
+      "step": 371
+    },
+    {
+      "epoch": 1.2200328407224958,
+      "grad_norm": 11.951933824951752,
+      "learning_rate": 7.894721963497213e-07,
+      "loss": 0.5991,
+      "step": 372
+    },
+    {
+      "epoch": 1.2233169129720853,
+      "grad_norm": 51.890411987411,
+      "learning_rate": 7.838713382712583e-07,
+      "loss": 0.5494,
+      "step": 373
+    },
+    {
+      "epoch": 1.2266009852216748,
+      "grad_norm": 46.51346153636678,
+      "learning_rate": 7.78277583323452e-07,
+      "loss": 0.5601,
+      "step": 374
+    },
+    {
+      "epoch": 1.2298850574712643,
+      "grad_norm": 32.19714846447316,
+      "learning_rate": 7.726911153466697e-07,
+      "loss": 0.5891,
+      "step": 375
+    },
+    {
+      "epoch": 1.2331691297208538,
+      "grad_norm": 30.416108844072596,
+      "learning_rate": 7.671121179417879e-07,
+      "loss": 0.57,
+      "step": 376
+    },
+    {
+      "epoch": 1.2364532019704433,
+      "grad_norm": 9.175156240446686,
+      "learning_rate": 7.615407744641618e-07,
+      "loss": 0.592,
+      "step": 377
+    },
+    {
+      "epoch": 1.2397372742200328,
+      "grad_norm": 16.89408306679232,
+      "learning_rate": 7.559772680175978e-07,
+      "loss": 0.5581,
+      "step": 378
+    },
+    {
+      "epoch": 1.2430213464696223,
+      "grad_norm": 36.581208571354075,
+      "learning_rate": 7.504217814483363e-07,
+      "loss": 0.5684,
+      "step": 379
+    },
+    {
+      "epoch": 1.2463054187192117,
+      "grad_norm": 197.1580272004894,
+      "learning_rate": 7.448744973390422e-07,
+      "loss": 0.5843,
+      "step": 380
+    },
+    {
+      "epoch": 1.2463054187192117,
+      "eval_loss": 0.3425952196121216,
+      "eval_runtime": 254.289,
+      "eval_samples_per_second": 8.125,
+      "eval_steps_per_second": 0.256,
+      "step": 380
+    },
+    {
+      "epoch": 1.2495894909688012,
+      "grad_norm": 38.00508215291586,
+      "learning_rate": 7.393355980028038e-07,
+      "loss": 0.5705,
+      "step": 381
+    },
+    {
+      "epoch": 1.2528735632183907,
+      "grad_norm": 36.66893410967853,
+      "learning_rate": 7.338052654771407e-07,
+      "loss": 0.5621,
+      "step": 382
+    },
+    {
+      "epoch": 1.2561576354679804,
+      "grad_norm": 72.0846419386083,
+      "learning_rate": 7.28283681518024e-07,
+      "loss": 0.6053,
+      "step": 383
+    },
+    {
+      "epoch": 1.2594417077175697,
+      "grad_norm": 12.87863278880059,
+      "learning_rate": 7.227710275938987e-07,
+      "loss": 0.585,
+      "step": 384
+    },
+    {
+      "epoch": 1.2627257799671594,
+      "grad_norm": 13.782452001612908,
+      "learning_rate": 7.172674848797217e-07,
+      "loss": 0.5973,
+      "step": 385
+    },
+    {
+      "epoch": 1.2660098522167487,
+      "grad_norm": 208.71637259406816,
+      "learning_rate": 7.117732342510092e-07,
+      "loss": 0.586,
+      "step": 386
+    },
+    {
+      "epoch": 1.2692939244663384,
+      "grad_norm": 16.912304799585513,
+      "learning_rate": 7.062884562778882e-07,
+      "loss": 0.5636,
+      "step": 387
+    },
+    {
+      "epoch": 1.2725779967159276,
+      "grad_norm": 28.08532516731275,
+      "learning_rate": 7.008133312191649e-07,
+      "loss": 0.5897,
+      "step": 388
+    },
+    {
+      "epoch": 1.2758620689655173,
+      "grad_norm": 55.25592778507127,
+      "learning_rate": 6.953480390164e-07,
+      "loss": 0.5581,
+      "step": 389
+    },
+    {
+      "epoch": 1.2791461412151066,
+      "grad_norm": 52.471536677355836,
+      "learning_rate": 6.898927592879944e-07,
+      "loss": 0.5488,
+      "step": 390
+    },
+    {
+      "epoch": 1.2824302134646963,
+      "grad_norm": 33.60924814778164,
+      "learning_rate": 6.844476713232862e-07,
+      "loss": 0.5452,
+      "step": 391
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 44.60405738816861,
+      "learning_rate": 6.79012954076658e-07,
+      "loss": 0.6057,
+      "step": 392
+    },
+    {
+      "epoch": 1.2889983579638753,
+      "grad_norm": 52.13213673114732,
+      "learning_rate": 6.735887861616555e-07,
+      "loss": 0.5718,
+      "step": 393
+    },
+    {
+      "epoch": 1.2922824302134646,
+      "grad_norm": 15.614998167223849,
+      "learning_rate": 6.681753458451189e-07,
+      "loss": 0.6077,
+      "step": 394
+    },
+    {
+      "epoch": 1.2955665024630543,
+      "grad_norm": 21.557430927559825,
+      "learning_rate": 6.627728110413213e-07,
+      "loss": 0.5724,
+      "step": 395
+    },
+    {
+      "epoch": 1.2988505747126438,
+      "grad_norm": 23.88672470819002,
+      "learning_rate": 6.573813593061235e-07,
+      "loss": 0.5623,
+      "step": 396
+    },
+    {
+      "epoch": 1.3021346469622332,
+      "grad_norm": 95.59341827721282,
+      "learning_rate": 6.520011678311381e-07,
+      "loss": 0.5932,
+      "step": 397
+    },
+    {
+      "epoch": 1.3054187192118227,
+      "grad_norm": 44.267252382206344,
+      "learning_rate": 6.466324134379065e-07,
+      "loss": 0.5699,
+      "step": 398
+    },
+    {
+      "epoch": 1.3087027914614122,
+      "grad_norm": 20.628188906562613,
+      "learning_rate": 6.412752725720864e-07,
+      "loss": 0.5971,
+      "step": 399
+    },
+    {
+      "epoch": 1.3119868637110017,
+      "grad_norm": 16.82849216948797,
+      "learning_rate": 6.359299212976534e-07,
+      "loss": 0.5843,
+      "step": 400
+    },
+    {
+      "epoch": 1.3152709359605912,
+      "grad_norm": 18.08566660075251,
+      "learning_rate": 6.305965352911161e-07,
+      "loss": 0.5404,
+      "step": 401
+    },
+    {
+      "epoch": 1.3185550082101807,
+      "grad_norm": 49.91580760003982,
+      "learning_rate": 6.252752898357397e-07,
+      "loss": 0.5531,
+      "step": 402
+    },
+    {
+      "epoch": 1.3218390804597702,
+      "grad_norm": 15.520146318832731,
+      "learning_rate": 6.199663598157875e-07,
+      "loss": 0.5872,
+      "step": 403
+    },
+    {
+      "epoch": 1.3251231527093597,
+      "grad_norm": 14.249014867996666,
+      "learning_rate": 6.146699197107715e-07,
+      "loss": 0.5623,
+      "step": 404
+    },
+    {
+      "epoch": 1.3284072249589491,
+      "grad_norm": 13.497082574794579,
+      "learning_rate": 6.093861435897207e-07,
+      "loss": 0.5707,
+      "step": 405
+    },
+    {
+      "epoch": 1.3316912972085386,
+      "grad_norm": 26.344677464265622,
+      "learning_rate": 6.041152051054575e-07,
+      "loss": 0.5597,
+      "step": 406
+    },
+    {
+      "epoch": 1.3349753694581281,
+      "grad_norm": 14.044122777177984,
+      "learning_rate": 5.988572774888912e-07,
+      "loss": 0.5924,
+      "step": 407
+    },
+    {
+      "epoch": 1.3382594417077176,
+      "grad_norm": 52.51863026094184,
+      "learning_rate": 5.936125335433264e-07,
+      "loss": 0.5774,
+      "step": 408
+    },
+    {
+      "epoch": 1.341543513957307,
+      "grad_norm": 36.13736458430556,
+      "learning_rate": 5.88381145638782e-07,
+      "loss": 0.5676,
+      "step": 409
+    },
+    {
+      "epoch": 1.3448275862068966,
+      "grad_norm": 15.292686883515266,
+      "learning_rate": 5.83163285706327e-07,
+      "loss": 0.5533,
+      "step": 410
+    },
+    {
+      "epoch": 1.348111658456486,
+      "grad_norm": 87.55364165940539,
+      "learning_rate": 5.779591252324286e-07,
+      "loss": 0.5676,
+      "step": 411
+    },
+    {
+      "epoch": 1.3513957307060755,
+      "grad_norm": 90.41445749482193,
+      "learning_rate": 5.72768835253319e-07,
+      "loss": 0.6106,
+      "step": 412
+    },
+    {
+      "epoch": 1.354679802955665,
+      "grad_norm": 75.68880847383154,
+      "learning_rate": 5.67592586349372e-07,
+      "loss": 0.5798,
+      "step": 413
+    },
+    {
+      "epoch": 1.3579638752052545,
+      "grad_norm": 24.386055319065594,
+      "learning_rate": 5.624305486394967e-07,
+      "loss": 0.5831,
+      "step": 414
+    },
+    {
+      "epoch": 1.361247947454844,
+      "grad_norm": 218.5561635346433,
+      "learning_rate": 5.57282891775548e-07,
+      "loss": 0.5958,
+      "step": 415
+    },
+    {
+      "epoch": 1.3645320197044335,
+      "grad_norm": 38.47004577556696,
+      "learning_rate": 5.5214978493675e-07,
+      "loss": 0.5522,
+      "step": 416
+    },
+    {
+      "epoch": 1.367816091954023,
+      "grad_norm": 11.116447526825615,
+      "learning_rate": 5.470313968241358e-07,
+      "loss": 0.5655,
+      "step": 417
+    },
+    {
+      "epoch": 1.3711001642036125,
+      "grad_norm": 33.49087738887472,
+      "learning_rate": 5.419278956550036e-07,
+      "loss": 0.5818,
+      "step": 418
+    },
+    {
+      "epoch": 1.374384236453202,
+      "grad_norm": 45.327655233207565,
+      "learning_rate": 5.368394491573876e-07,
+      "loss": 0.5618,
+      "step": 419
+    },
+    {
+      "epoch": 1.3776683087027914,
+      "grad_norm": 266.5398840292457,
+      "learning_rate": 5.317662245645469e-07,
+      "loss": 0.5549,
+      "step": 420
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 20.553987754267183,
+      "learning_rate": 5.267083886094668e-07,
+      "loss": 0.5719,
+      "step": 421
+    },
+    {
+      "epoch": 1.3842364532019704,
+      "grad_norm": 74.60343344764495,
+      "learning_rate": 5.216661075193813e-07,
+      "loss": 0.5709,
+      "step": 422
+    },
+    {
+      "epoch": 1.38752052545156,
+      "grad_norm": 81.24452905189378,
+      "learning_rate": 5.166395470103091e-07,
+      "loss": 0.5839,
+      "step": 423
+    },
+    {
+      "epoch": 1.3908045977011494,
+      "grad_norm": 61.57043976405265,
+      "learning_rate": 5.116288722816085e-07,
+      "loss": 0.5881,
+      "step": 424
+    },
+    {
+      "epoch": 1.3940886699507389,
+      "grad_norm": 66.61917309787378,
+      "learning_rate": 5.066342480105459e-07,
+      "loss": 0.5396,
+      "step": 425
+    },
+    {
+      "epoch": 1.3973727422003284,
+      "grad_norm": 32.18086853290272,
+      "learning_rate": 5.016558383468851e-07,
+      "loss": 0.5743,
+      "step": 426
+    },
+    {
+      "epoch": 1.4006568144499179,
+      "grad_norm": 20.082111947795827,
+      "learning_rate": 4.966938069074921e-07,
+      "loss": 0.5531,
+      "step": 427
+    },
+    {
+      "epoch": 1.4039408866995073,
+      "grad_norm": 35.324879089856665,
+      "learning_rate": 4.91748316770958e-07,
+      "loss": 0.5685,
+      "step": 428
+    },
+    {
+      "epoch": 1.4072249589490968,
+      "grad_norm": 52.60744635091702,
+      "learning_rate": 4.868195304722391e-07,
+      "loss": 0.5892,
+      "step": 429
+    },
+    {
+      "epoch": 1.4105090311986863,
+      "grad_norm": 32.96054620176144,
+      "learning_rate": 4.819076099973152e-07,
+      "loss": 0.5593,
+      "step": 430
+    },
+    {
+      "epoch": 1.4137931034482758,
+      "grad_norm": 203.82839428938274,
+      "learning_rate": 4.770127167778669e-07,
+      "loss": 0.5699,
+      "step": 431
+    },
+    {
+      "epoch": 1.4170771756978653,
+      "grad_norm": 47.57730637955671,
+      "learning_rate": 4.7213501168596746e-07,
+      "loss": 0.5789,
+      "step": 432
+    },
+    {
+      "epoch": 1.4203612479474548,
+      "grad_norm": 67.44290592976088,
+      "learning_rate": 4.6727465502879846e-07,
+      "loss": 0.5556,
+      "step": 433
+    },
+    {
+      "epoch": 1.4236453201970443,
+      "grad_norm": 44.034037450189444,
+      "learning_rate": 4.6243180654337966e-07,
+      "loss": 0.5549,
+      "step": 434
+    },
+    {
+      "epoch": 1.4269293924466337,
+      "grad_norm": 16.61985345751509,
+      "learning_rate": 4.5760662539132077e-07,
+      "loss": 0.5747,
+      "step": 435
+    },
+    {
+      "epoch": 1.4302134646962232,
+      "grad_norm": 12.882168633818488,
+      "learning_rate": 4.5279927015358833e-07,
+      "loss": 0.5717,
+      "step": 436
+    },
+    {
+      "epoch": 1.4334975369458127,
+      "grad_norm": 20.653148975087017,
+      "learning_rate": 4.480098988252957e-07,
+      "loss": 0.6128,
+      "step": 437
+    },
+    {
+      "epoch": 1.4367816091954024,
+      "grad_norm": 23.18939767829898,
+      "learning_rate": 4.4323866881050945e-07,
+      "loss": 0.5692,
+      "step": 438
+    },
+    {
+      "epoch": 1.4400656814449917,
+      "grad_norm": 31.061046848137774,
+      "learning_rate": 4.384857369170771e-07,
+      "loss": 0.5611,
+      "step": 439
+    },
+    {
+      "epoch": 1.4433497536945814,
+      "grad_norm": 62.563561066656575,
+      "learning_rate": 4.337512593514728e-07,
+      "loss": 0.56,
+      "step": 440
+    },
+    {
+      "epoch": 1.4466338259441707,
+      "grad_norm": 17.20298230988248,
+      "learning_rate": 4.290353917136639e-07,
+      "loss": 0.5496,
+      "step": 441
+    },
+    {
+      "epoch": 1.4499178981937604,
+      "grad_norm": 10.022400404488588,
+      "learning_rate": 4.2433828899199807e-07,
+      "loss": 0.5743,
+      "step": 442
+    },
+    {
+      "epoch": 1.4532019704433496,
+      "grad_norm": 29.312045763415277,
+      "learning_rate": 4.1966010555810694e-07,
+      "loss": 0.5445,
+      "step": 443
+    },
+    {
+      "epoch": 1.4564860426929394,
+      "grad_norm": 31.664457234542283,
+      "learning_rate": 4.150009951618355e-07,
+      "loss": 0.55,
+      "step": 444
+    },
+    {
+      "epoch": 1.4597701149425286,
+      "grad_norm": 23.68782329905495,
+      "learning_rate": 4.103611109261872e-07,
+      "loss": 0.5957,
+      "step": 445
+    },
+    {
+      "epoch": 1.4630541871921183,
+      "grad_norm": 15.601252467779627,
+      "learning_rate": 4.0574060534229327e-07,
+      "loss": 0.5965,
+      "step": 446
+    },
+    {
+      "epoch": 1.4663382594417076,
+      "grad_norm": 16.094218846568037,
+      "learning_rate": 4.011396302643988e-07,
+      "loss": 0.5731,
+      "step": 447
+    },
+    {
+      "epoch": 1.4696223316912973,
+      "grad_norm": 45.95169139269504,
+      "learning_rate": 3.965583369048737e-07,
+      "loss": 0.5778,
+      "step": 448
+    },
+    {
+      "epoch": 1.4729064039408866,
+      "grad_norm": 16.32693066785533,
+      "learning_rate": 3.9199687582924246e-07,
+      "loss": 0.582,
+      "step": 449
+    },
+    {
+      "epoch": 1.4761904761904763,
+      "grad_norm": 15.3081339931233,
+      "learning_rate": 3.8745539695123577e-07,
+      "loss": 0.5865,
+      "step": 450
+    },
+    {
+      "epoch": 1.4794745484400658,
+      "grad_norm": 41.26715984492636,
+      "learning_rate": 3.829340495278639e-07,
+      "loss": 0.5671,
+      "step": 451
+    },
+    {
+      "epoch": 1.4827586206896552,
+      "grad_norm": 14.84205956050775,
+      "learning_rate": 3.7843298215451046e-07,
+      "loss": 0.5676,
+      "step": 452
+    },
+    {
+      "epoch": 1.4860426929392447,
+      "grad_norm": 28.834583020994884,
+      "learning_rate": 3.739523427600508e-07,
+      "loss": 0.5831,
+      "step": 453
+    },
+    {
+      "epoch": 1.4893267651888342,
+      "grad_norm": 29.206427309967527,
+      "learning_rate": 3.6949227860198707e-07,
+      "loss": 0.5808,
+      "step": 454
+    },
+    {
+      "epoch": 1.4926108374384237,
+      "grad_norm": 21.208424115794667,
+      "learning_rate": 3.6505293626161127e-07,
+      "loss": 0.5701,
+      "step": 455
+    },
+    {
+      "epoch": 1.4958949096880132,
+      "grad_norm": 60.972107883689446,
+      "learning_rate": 3.6063446163918663e-07,
+      "loss": 0.5812,
+      "step": 456
+    },
+    {
+      "epoch": 1.4958949096880132,
+      "eval_loss": 0.34147635102272034,
+      "eval_runtime": 253.4458,
+      "eval_samples_per_second": 8.152,
+      "eval_steps_per_second": 0.256,
+      "step": 456
+    },
+    {
+      "epoch": 1.4991789819376027,
+      "grad_norm": 102.55339346012296,
+      "learning_rate": 3.5623699994915355e-07,
+      "loss": 0.5955,
+      "step": 457
+    },
+    {
+      "epoch": 1.5024630541871922,
+      "grad_norm": 76.36859833115236,
+      "learning_rate": 3.518606957153557e-07,
+      "loss": 0.5682,
+      "step": 458
+    },
+    {
+      "epoch": 1.5057471264367817,
+      "grad_norm": 24.75361464219726,
+      "learning_rate": 3.475056927662912e-07,
+      "loss": 0.5787,
+      "step": 459
+    },
+    {
+      "epoch": 1.5090311986863711,
+      "grad_norm": 27.83323345966019,
+      "learning_rate": 3.4317213423038384e-07,
+      "loss": 0.562,
+      "step": 460
+    },
+    {
+      "epoch": 1.5123152709359606,
+      "grad_norm": 46.505159610329876,
+      "learning_rate": 3.3886016253128327e-07,
+      "loss": 0.5851,
+      "step": 461
+    },
+    {
+      "epoch": 1.5155993431855501,
+      "grad_norm": 28.610228426478272,
+      "learning_rate": 3.345699193831795e-07,
+      "loss": 0.6063,
+      "step": 462
+    },
+    {
+      "epoch": 1.5188834154351396,
+      "grad_norm": 15.024413491910803,
+      "learning_rate": 3.303015457861478e-07,
+      "loss": 0.5277,
+      "step": 463
+    },
+    {
+      "epoch": 1.522167487684729,
+      "grad_norm": 12.258290893394133,
+      "learning_rate": 3.2605518202151574e-07,
+      "loss": 0.5922,
+      "step": 464
+    },
+    {
+      "epoch": 1.5254515599343186,
+      "grad_norm": 104.31924222325128,
+      "learning_rate": 3.2183096764724914e-07,
+      "loss": 0.5649,
+      "step": 465
+    },
+    {
+      "epoch": 1.528735632183908,
+      "grad_norm": 42.86381800387519,
+      "learning_rate": 3.1762904149336946e-07,
+      "loss": 0.5743,
+      "step": 466
+    },
+    {
+      "epoch": 1.5320197044334976,
+      "grad_norm": 18.167710389726654,
+      "learning_rate": 3.134495416573883e-07,
+      "loss": 0.5572,
+      "step": 467
+    },
+    {
+      "epoch": 1.535303776683087,
+      "grad_norm": 15.343486359801176,
+      "learning_rate": 3.092926054997711e-07,
+      "loss": 0.5719,
+      "step": 468
+    },
+    {
+      "epoch": 1.5385878489326765,
+      "grad_norm": 25.820050498948465,
+      "learning_rate": 3.0515836963942054e-07,
+      "loss": 0.5785,
+      "step": 469
+    },
+    {
+      "epoch": 1.541871921182266,
+      "grad_norm": 47.76654830827422,
+      "learning_rate": 3.01046969949188e-07,
+      "loss": 0.571,
+      "step": 470
+    },
+    {
+      "epoch": 1.5451559934318555,
+      "grad_norm": 18.685300965930125,
+      "learning_rate": 2.969585415514064e-07,
+      "loss": 0.5899,
+      "step": 471
+    },
+    {
+      "epoch": 1.548440065681445,
+      "grad_norm": 33.45206718151848,
+      "learning_rate": 2.9289321881345254e-07,
+      "loss": 0.5661,
+      "step": 472
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 72.16443965311746,
+      "learning_rate": 2.8885113534332737e-07,
+      "loss": 0.5931,
+      "step": 473
+    },
+    {
+      "epoch": 1.555008210180624,
+      "grad_norm": 42.035242935885876,
+      "learning_rate": 2.848324239852672e-07,
+      "loss": 0.5587,
+      "step": 474
+    },
+    {
+      "epoch": 1.5582922824302134,
+      "grad_norm": 45.15146887717905,
+      "learning_rate": 2.80837216815378e-07,
+      "loss": 0.5728,
+      "step": 475
+    },
+    {
+      "epoch": 1.561576354679803,
+      "grad_norm": 21.81085786153228,
+      "learning_rate": 2.768656451372919e-07,
+      "loss": 0.6011,
+      "step": 476
+    },
+    {
+      "epoch": 1.5648604269293924,
+      "grad_norm": 14.491067969072652,
+      "learning_rate": 2.729178394778554e-07,
+      "loss": 0.5558,
+      "step": 477
+    },
+    {
+      "epoch": 1.568144499178982,
+      "grad_norm": 17.912743886068196,
+      "learning_rate": 2.6899392958283706e-07,
+      "loss": 0.5631,
+      "step": 478
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 64.4857516790575,
+      "learning_rate": 2.6509404441266535e-07,
+      "loss": 0.5841,
+      "step": 479
+    },
+    {
+      "epoch": 1.5747126436781609,
+      "grad_norm": 40.36341697674323,
+      "learning_rate": 2.6121831213818826e-07,
+      "loss": 0.5328,
+      "step": 480
+    },
+    {
+      "epoch": 1.5779967159277504,
+      "grad_norm": 28.64813563751768,
+      "learning_rate": 2.573668601364623e-07,
+      "loss": 0.5665,
+      "step": 481
+    },
+    {
+      "epoch": 1.5812807881773399,
+      "grad_norm": 20.644096206862233,
+      "learning_rate": 2.5353981498656505e-07,
+      "loss": 0.5579,
+      "step": 482
+    },
+    {
+      "epoch": 1.5845648604269293,
+      "grad_norm": 42.66803007868702,
+      "learning_rate": 2.497373024654373e-07,
+      "loss": 0.5692,
+      "step": 483
+    },
+    {
+      "epoch": 1.5878489326765188,
+      "grad_norm": 23.741999865754043,
+      "learning_rate": 2.459594475437472e-07,
+      "loss": 0.5731,
+      "step": 484
+    },
+    {
+      "epoch": 1.5911330049261085,
+      "grad_norm": 25.83928960154385,
+      "learning_rate": 2.4220637438178313e-07,
+      "loss": 0.5895,
+      "step": 485
+    },
+    {
+      "epoch": 1.5944170771756978,
+      "grad_norm": 53.46424769894938,
+      "learning_rate": 2.3847820632537564e-07,
+      "loss": 0.5807,
+      "step": 486
+    },
+    {
+      "epoch": 1.5977011494252875,
+      "grad_norm": 32.58548532914844,
+      "learning_rate": 2.3477506590183972e-07,
+      "loss": 0.5525,
+      "step": 487
+    },
+    {
+      "epoch": 1.6009852216748768,
+      "grad_norm": 40.67612047783848,
+      "learning_rate": 2.310970748159511e-07,
+      "loss": 0.6111,
+      "step": 488
+    },
+    {
+      "epoch": 1.6042692939244665,
+      "grad_norm": 31.120710466672246,
+      "learning_rate": 2.2744435394594497e-07,
+      "loss": 0.569,
+      "step": 489
+    },
+    {
+      "epoch": 1.6075533661740558,
+      "grad_norm": 11.431716307911014,
+      "learning_rate": 2.2381702333954433e-07,
+      "loss": 0.5804,
+      "step": 490
+    },
+    {
+      "epoch": 1.6108374384236455,
+      "grad_norm": 13.840123564647127,
+      "learning_rate": 2.2021520221001299e-07,
+      "loss": 0.57,
+      "step": 491
+    },
+    {
+      "epoch": 1.6141215106732347,
+      "grad_norm": 17.928907652041175,
+      "learning_rate": 2.1663900893223897e-07,
+      "loss": 0.5922,
+      "step": 492
+    },
+    {
+      "epoch": 1.6174055829228244,
+      "grad_norm": 57.74383968458873,
+      "learning_rate": 2.1308856103884277e-07,
+      "loss": 0.5681,
+      "step": 493
+    },
+    {
+      "epoch": 1.6206896551724137,
+      "grad_norm": 49.41941546684546,
+      "learning_rate": 2.0956397521631662e-07,
+      "loss": 0.5555,
+      "step": 494
+    },
+    {
+      "epoch": 1.6239737274220034,
+      "grad_norm": 30.887246526627205,
+      "learning_rate": 2.0606536730118763e-07,
+      "loss": 0.5666,
+      "step": 495
+    },
+    {
+      "epoch": 1.6272577996715927,
+      "grad_norm": 21.706062838691153,
+      "learning_rate": 2.0259285227621147e-07,
+      "loss": 0.5633,
+      "step": 496
+    },
+    {
+      "epoch": 1.6305418719211824,
+      "grad_norm": 31.266696706985737,
+      "learning_rate": 1.991465442665937e-07,
+      "loss": 0.5356,
+      "step": 497
+    },
+    {
+      "epoch": 1.6338259441707716,
+      "grad_norm": 29.3689515647996,
+      "learning_rate": 1.9572655653623882e-07,
+      "loss": 0.5912,
+      "step": 498
+    },
+    {
+      "epoch": 1.6371100164203614,
+      "grad_norm": 22.322656636658127,
+      "learning_rate": 1.9233300148402764e-07,
+      "loss": 0.5595,
+      "step": 499
+    },
+    {
+      "epoch": 1.6403940886699506,
+      "grad_norm": 44.171884336035255,
+      "learning_rate": 1.8896599064012298e-07,
+      "loss": 0.5643,
+      "step": 500
+    },
+    {
+      "epoch": 1.6436781609195403,
+      "grad_norm": 17.948977196325135,
+      "learning_rate": 1.8562563466230575e-07,
+      "loss": 0.5634,
+      "step": 501
+    },
+    {
+      "epoch": 1.6469622331691296,
+      "grad_norm": 11.97820725055819,
+      "learning_rate": 1.8231204333233607e-07,
+      "loss": 0.5598,
+      "step": 502
+    },
+    {
+      "epoch": 1.6502463054187193,
+      "grad_norm": 10.612778968035474,
+      "learning_rate": 1.790253255523465e-07,
+      "loss": 0.5803,
+      "step": 503
+    },
+    {
+      "epoch": 1.6535303776683086,
+      "grad_norm": 18.193226588472413,
+      "learning_rate": 1.7576558934126217e-07,
+      "loss": 0.5967,
+      "step": 504
+    },
+    {
+      "epoch": 1.6568144499178983,
+      "grad_norm": 12.00030723839093,
+      "learning_rate": 1.7253294183125222e-07,
+      "loss": 0.5796,
+      "step": 505
+    },
+    {
+      "epoch": 1.6600985221674875,
+      "grad_norm": 61.889660413482744,
+      "learning_rate": 1.6932748926420693e-07,
+      "loss": 0.6197,
+      "step": 506
+    },
+    {
+      "epoch": 1.6633825944170773,
+      "grad_norm": 14.289243326025748,
+      "learning_rate": 1.6614933698824728e-07,
+      "loss": 0.5715,
+      "step": 507
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 83.1922121560888,
+      "learning_rate": 1.6299858945426248e-07,
+      "loss": 0.6025,
+      "step": 508
+    },
+    {
+      "epoch": 1.6699507389162562,
+      "grad_norm": 21.289795508971476,
+      "learning_rate": 1.5987535021247667e-07,
+      "loss": 0.5613,
+      "step": 509
+    },
+    {
+      "epoch": 1.6732348111658455,
+      "grad_norm": 9.12840521329095,
+      "learning_rate": 1.5677972190904621e-07,
+      "loss": 0.5611,
+      "step": 510
+    },
+    {
+      "epoch": 1.6765188834154352,
+      "grad_norm": 80.54244783118818,
+      "learning_rate": 1.5371180628268587e-07,
+      "loss": 0.5773,
+      "step": 511
+    },
+    {
+      "epoch": 1.6798029556650245,
+      "grad_norm": 18.09168094972728,
+      "learning_rate": 1.5067170416132603e-07,
+      "loss": 0.5632,
+      "step": 512
+    },
+    {
+      "epoch": 1.6830870279146142,
+      "grad_norm": 76.91613553570883,
+      "learning_rate": 1.476595154587973e-07,
+      "loss": 0.5859,
+      "step": 513
+    },
+    {
+      "epoch": 1.6863711001642037,
+      "grad_norm": 13.09746011495923,
+      "learning_rate": 1.446753391715484e-07,
+      "loss": 0.5555,
+      "step": 514
+    },
+    {
+      "epoch": 1.6896551724137931,
+      "grad_norm": 26.40133497602893,
+      "learning_rate": 1.4171927337539104e-07,
+      "loss": 0.5222,
+      "step": 515
+    },
+    {
+      "epoch": 1.6929392446633826,
+      "grad_norm": 42.003291635263835,
+      "learning_rate": 1.3879141522227877e-07,
+      "loss": 0.5857,
+      "step": 516
+    },
+    {
+      "epoch": 1.6962233169129721,
+      "grad_norm": 15.138211036731974,
+      "learning_rate": 1.3589186093711223e-07,
+      "loss": 0.576,
+      "step": 517
+    },
+    {
+      "epoch": 1.6995073891625616,
+      "grad_norm": 15.749783656567669,
+      "learning_rate": 1.3302070581457713e-07,
+      "loss": 0.5616,
+      "step": 518
+    },
+    {
+      "epoch": 1.702791461412151,
+      "grad_norm": 19.998784663850575,
+      "learning_rate": 1.3017804421601298e-07,
+      "loss": 0.6002,
+      "step": 519
+    },
+    {
+      "epoch": 1.7060755336617406,
+      "grad_norm": 12.84175073867242,
+      "learning_rate": 1.273639695663108e-07,
+      "loss": 0.5693,
+      "step": 520
+    },
+    {
+      "epoch": 1.70935960591133,
+      "grad_norm": 48.66516503978152,
+      "learning_rate": 1.2457857435084407e-07,
+      "loss": 0.6153,
+      "step": 521
+    },
+    {
+      "epoch": 1.7126436781609196,
+      "grad_norm": 16.696763387794824,
+      "learning_rate": 1.2182195011242747e-07,
+      "loss": 0.5613,
+      "step": 522
+    },
+    {
+      "epoch": 1.715927750410509,
+      "grad_norm": 31.172316224112276,
+      "learning_rate": 1.1909418744831046e-07,
+      "loss": 0.605,
+      "step": 523
+    },
+    {
+      "epoch": 1.7192118226600985,
+      "grad_norm": 19.082497779204967,
+      "learning_rate": 1.1639537600719761e-07,
+      "loss": 0.5805,
+      "step": 524
+    },
+    {
+      "epoch": 1.722495894909688,
+      "grad_norm": 15.143716837248418,
+      "learning_rate": 1.1372560448630375e-07,
+      "loss": 0.608,
+      "step": 525
+    },
+    {
+      "epoch": 1.7257799671592775,
+      "grad_norm": 37.76170562223439,
+      "learning_rate": 1.1108496062843741e-07,
+      "loss": 0.5838,
+      "step": 526
+    },
+    {
+      "epoch": 1.729064039408867,
+      "grad_norm": 14.565254764998345,
+      "learning_rate": 1.0847353121911951e-07,
+      "loss": 0.5539,
+      "step": 527
+    },
+    {
+      "epoch": 1.7323481116584565,
+      "grad_norm": 32.47683113493387,
+      "learning_rate": 1.0589140208372871e-07,
+      "loss": 0.5575,
+      "step": 528
+    },
+    {
+      "epoch": 1.735632183908046,
+      "grad_norm": 23.18347373384476,
+      "learning_rate": 1.0333865808468201e-07,
+      "loss": 0.5566,
+      "step": 529
+    },
+    {
+      "epoch": 1.7389162561576355,
+      "grad_norm": 22.4422071324967,
+      "learning_rate": 1.0081538311864568e-07,
+      "loss": 0.576,
+      "step": 530
+    },
+    {
+      "epoch": 1.742200328407225,
+      "grad_norm": 14.059292999171634,
+      "learning_rate": 9.83216601137773e-08,
+      "loss": 0.5677,
+      "step": 531
+    },
+    {
+      "epoch": 1.7454844006568144,
+      "grad_norm": 22.050768611070254,
+      "learning_rate": 9.58575710270011e-08,
+      "loss": 0.5676,
+      "step": 532
+    },
+    {
+      "epoch": 1.7454844006568144,
+      "eval_loss": 0.34110337495803833,
+      "eval_runtime": 255.1137,
+      "eval_samples_per_second": 8.098,
+      "eval_steps_per_second": 0.255,
+      "step": 532
+    },
+    {
+      "epoch": 1.748768472906404,
+      "grad_norm": 28.327968167510832,
+      "learning_rate": 9.342319684131395e-08,
+      "loss": 0.553,
+      "step": 533
+    },
+    {
+      "epoch": 1.7520525451559934,
+      "grad_norm": 23.765559507553224,
+      "learning_rate": 9.101861756312368e-08,
+      "loss": 0.5751,
+      "step": 534
+    },
+    {
+      "epoch": 1.7553366174055829,
+      "grad_norm": 38.32793241416281,
+      "learning_rate": 8.864391221962064e-08,
+      "loss": 0.5721,
+      "step": 535
+    },
+    {
+      "epoch": 1.7586206896551724,
+      "grad_norm": 23.1435189905778,
+      "learning_rate": 8.62991588561791e-08,
+      "loss": 0.5372,
+      "step": 536
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 64.41410656507954,
+      "learning_rate": 8.398443453379266e-08,
+      "loss": 0.597,
+      "step": 537
+    },
+    {
+      "epoch": 1.7651888341543513,
+      "grad_norm": 26.570215615082457,
+      "learning_rate": 8.169981532654269e-08,
+      "loss": 0.5425,
+      "step": 538
+    },
+    {
+      "epoch": 1.7684729064039408,
+      "grad_norm": 19.853728565908686,
+      "learning_rate": 7.944537631909664e-08,
+      "loss": 0.5781,
+      "step": 539
+    },
+    {
+      "epoch": 1.7717569786535303,
+      "grad_norm": 40.09614615765141,
+      "learning_rate": 7.722119160424112e-08,
+      "loss": 0.5738,
+      "step": 540
+    },
+    {
+      "epoch": 1.7750410509031198,
+      "grad_norm": 56.37553140710831,
+      "learning_rate": 7.502733428044683e-08,
+      "loss": 0.5788,
+      "step": 541
+    },
+    {
+      "epoch": 1.7783251231527095,
+      "grad_norm": 25.79428871095927,
+      "learning_rate": 7.286387644946601e-08,
+      "loss": 0.5614,
+      "step": 542
+    },
+    {
+      "epoch": 1.7816091954022988,
+      "grad_norm": 37.57206986640528,
+      "learning_rate": 7.073088921396286e-08,
+      "loss": 0.6319,
+      "step": 543
+    },
+    {
+      "epoch": 1.7848932676518885,
+      "grad_norm": 78.07770173709457,
+      "learning_rate": 6.862844267517642e-08,
+      "loss": 0.5585,
+      "step": 544
+    },
+    {
+      "epoch": 1.7881773399014778,
+      "grad_norm": 45.27512632946113,
+      "learning_rate": 6.655660593061718e-08,
+      "loss": 0.5835,
+      "step": 545
+    },
+    {
+      "epoch": 1.7914614121510675,
+      "grad_norm": 43.444015827492656,
+      "learning_rate": 6.451544707179635e-08,
+      "loss": 0.5839,
+      "step": 546
+    },
+    {
+      "epoch": 1.7947454844006567,
+      "grad_norm": 30.129718093780813,
+      "learning_rate": 6.250503318198663e-08,
+      "loss": 0.5922,
+      "step": 547
+    },
+    {
+      "epoch": 1.7980295566502464,
+      "grad_norm": 12.289775538527161,
+      "learning_rate": 6.052543033401891e-08,
+      "loss": 0.5659,
+      "step": 548
+    },
+    {
+      "epoch": 1.8013136288998357,
+      "grad_norm": 22.117467884787505,
+      "learning_rate": 5.8576703588110953e-08,
+      "loss": 0.5833,
+      "step": 549
+    },
+    {
+      "epoch": 1.8045977011494254,
+      "grad_norm": 18.824028673171902,
+      "learning_rate": 5.665891698972769e-08,
+      "loss": 0.5515,
+      "step": 550
+    },
+    {
+      "epoch": 1.8078817733990147,
+      "grad_norm": 15.156245839885349,
+      "learning_rate": 5.4772133567477454e-08,
+      "loss": 0.58,
+      "step": 551
+    },
+    {
+      "epoch": 1.8111658456486044,
+      "grad_norm": 17.302260811016588,
+      "learning_rate": 5.291641533104052e-08,
+      "loss": 0.5443,
+      "step": 552
+    },
+    {
+      "epoch": 1.8144499178981937,
+      "grad_norm": 72.23650640208795,
+      "learning_rate": 5.109182326913053e-08,
+      "loss": 0.5913,
+      "step": 553
+    },
+    {
+      "epoch": 1.8177339901477834,
+      "grad_norm": 41.839452624673,
+      "learning_rate": 4.929841734749063e-08,
+      "loss": 0.5826,
+      "step": 554
+    },
+    {
+      "epoch": 1.8210180623973726,
+      "grad_norm": 150.9975193542559,
+      "learning_rate": 4.75362565069225e-08,
+      "loss": 0.5672,
+      "step": 555
+    },
+    {
+      "epoch": 1.8243021346469623,
+      "grad_norm": 177.35572727759407,
+      "learning_rate": 4.580539866134914e-08,
+      "loss": 0.5962,
+      "step": 556
+    },
+    {
+      "epoch": 1.8275862068965516,
+      "grad_norm": 11.471174551449721,
+      "learning_rate": 4.410590069591191e-08,
+      "loss": 0.5568,
+      "step": 557
+    },
+    {
+      "epoch": 1.8308702791461413,
+      "grad_norm": 41.93391866191071,
+      "learning_rate": 4.2437818465100306e-08,
+      "loss": 0.5872,
+      "step": 558
+    },
+    {
+      "epoch": 1.8341543513957306,
+      "grad_norm": 49.54043761767362,
+      "learning_rate": 4.080120679091681e-08,
+      "loss": 0.5782,
+      "step": 559
+    },
+    {
+      "epoch": 1.8374384236453203,
+      "grad_norm": 43.81540298910706,
+      "learning_rate": 3.919611946107493e-08,
+      "loss": 0.5726,
+      "step": 560
+    },
+    {
+      "epoch": 1.8407224958949095,
+      "grad_norm": 23.38299802172784,
+      "learning_rate": 3.762260922723182e-08,
+      "loss": 0.5658,
+      "step": 561
+    },
+    {
+      "epoch": 1.8440065681444993,
+      "grad_norm": 24.53118761357852,
+      "learning_rate": 3.6080727803253996e-08,
+      "loss": 0.5715,
+      "step": 562
+    },
+    {
+      "epoch": 1.8472906403940885,
+      "grad_norm": 19.417907974207314,
+      "learning_rate": 3.4570525863518164e-08,
+      "loss": 0.591,
+      "step": 563
+    },
+    {
+      "epoch": 1.8505747126436782,
+      "grad_norm": 10.93698504246622,
+      "learning_rate": 3.309205304124552e-08,
+      "loss": 0.598,
+      "step": 564
+    },
+    {
+      "epoch": 1.8538587848932675,
+      "grad_norm": 14.732319976335372,
+      "learning_rate": 3.164535792687095e-08,
+      "loss": 0.5739,
+      "step": 565
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 17.119911911692302,
+      "learning_rate": 3.0230488066445457e-08,
+      "loss": 0.5932,
+      "step": 566
+    },
+    {
+      "epoch": 1.8604269293924465,
+      "grad_norm": 18.02807205348044,
+      "learning_rate": 2.8847489960074133e-08,
+      "loss": 0.5698,
+      "step": 567
+    },
+    {
+      "epoch": 1.8637110016420362,
+      "grad_norm": 42.32940645169373,
+      "learning_rate": 2.749640906038797e-08,
+      "loss": 0.5845,
+      "step": 568
+    },
+    {
+      "epoch": 1.8669950738916257,
+      "grad_norm": 21.291255161442834,
+      "learning_rate": 2.617728977104927e-08,
+      "loss": 0.5767,
+      "step": 569
+    },
+    {
+      "epoch": 1.8702791461412152,
+      "grad_norm": 126.41354720715246,
+      "learning_rate": 2.4890175445293147e-08,
+      "loss": 0.548,
+      "step": 570
+    },
+    {
+      "epoch": 1.8735632183908046,
+      "grad_norm": 89.21090716126868,
+      "learning_rate": 2.3635108384502e-08,
+      "loss": 0.5526,
+      "step": 571
+    },
+    {
+      "epoch": 1.8768472906403941,
+      "grad_norm": 28.90642350240842,
+      "learning_rate": 2.2412129836816285e-08,
+      "loss": 0.5923,
+      "step": 572
+    },
+    {
+      "epoch": 1.8801313628899836,
+      "grad_norm": 33.649895492267454,
+      "learning_rate": 2.122127999577783e-08,
+      "loss": 0.5915,
+      "step": 573
+    },
+    {
+      "epoch": 1.883415435139573,
+      "grad_norm": 42.351338285848115,
+      "learning_rate": 2.0062597999009114e-08,
+      "loss": 0.5692,
+      "step": 574
+    },
+    {
+      "epoch": 1.8866995073891626,
+      "grad_norm": 21.17999827332917,
+      "learning_rate": 1.8936121926927507e-08,
+      "loss": 0.5528,
+      "step": 575
+    },
+    {
+      "epoch": 1.889983579638752,
+      "grad_norm": 15.439339266974041,
+      "learning_rate": 1.7841888801493176e-08,
+      "loss": 0.5694,
+      "step": 576
+    },
+    {
+      "epoch": 1.8932676518883416,
+      "grad_norm": 13.470779589141687,
+      "learning_rate": 1.6779934584992718e-08,
+      "loss": 0.5879,
+      "step": 577
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 27.077399389940133,
+      "learning_rate": 1.575029417885687e-08,
+      "loss": 0.5641,
+      "step": 578
+    },
+    {
+      "epoch": 1.8998357963875205,
+      "grad_norm": 19.647395918482665,
+      "learning_rate": 1.4753001422514121e-08,
+      "loss": 0.5752,
+      "step": 579
+    },
+    {
+      "epoch": 1.90311986863711,
+      "grad_norm": 33.07929236692647,
+      "learning_rate": 1.3788089092277688e-08,
+      "loss": 0.5829,
+      "step": 580
+    },
+    {
+      "epoch": 1.9064039408866995,
+      "grad_norm": 80.07673176809891,
+      "learning_rate": 1.2855588900269054e-08,
+      "loss": 0.5497,
+      "step": 581
+    },
+    {
+      "epoch": 1.909688013136289,
+      "grad_norm": 19.376860505012647,
+      "learning_rate": 1.1955531493375138e-08,
+      "loss": 0.6026,
+      "step": 582
+    },
+    {
+      "epoch": 1.9129720853858785,
+      "grad_norm": 35.47816644500678,
+      "learning_rate": 1.108794645224187e-08,
+      "loss": 0.5666,
+      "step": 583
+    },
+    {
+      "epoch": 1.916256157635468,
+      "grad_norm": 115.1418329735163,
+      "learning_rate": 1.0252862290301089e-08,
+      "loss": 0.5753,
+      "step": 584
+    },
+    {
+      "epoch": 1.9195402298850575,
+      "grad_norm": 30.510225961444675,
+      "learning_rate": 9.450306452834178e-09,
+      "loss": 0.5848,
+      "step": 585
+    },
+    {
+      "epoch": 1.922824302134647,
+      "grad_norm": 46.46127331757118,
+      "learning_rate": 8.68030531606967e-09,
+      "loss": 0.5815,
+      "step": 586
+    },
+    {
+      "epoch": 1.9261083743842364,
+      "grad_norm": 21.7313503869492,
+      "learning_rate": 7.94288418631639e-09,
+      "loss": 0.5717,
+      "step": 587
+    },
+    {
+      "epoch": 1.929392446633826,
+      "grad_norm": 127.17771447820743,
+      "learning_rate": 7.2380672991319e-09,
+      "loss": 0.571,
+      "step": 588
+    },
+    {
+      "epoch": 1.9326765188834154,
+      "grad_norm": 31.316585311430142,
+      "learning_rate": 6.565877818526244e-09,
+      "loss": 0.5453,
+      "step": 589
+    },
+    {
+      "epoch": 1.935960591133005,
+      "grad_norm": 51.54565942906159,
+      "learning_rate": 5.926337836199891e-09,
+      "loss": 0.5669,
+      "step": 590
+    },
+    {
+      "epoch": 1.9392446633825944,
+      "grad_norm": 17.804113285989903,
+      "learning_rate": 5.319468370818536e-09,
+      "loss": 0.601,
+      "step": 591
+    },
+    {
+      "epoch": 1.9425287356321839,
+      "grad_norm": 68.75929656403164,
+      "learning_rate": 4.745289367321658e-09,
+      "loss": 0.5976,
+      "step": 592
+    },
+    {
+      "epoch": 1.9458128078817734,
+      "grad_norm": 11.91134893164374,
+      "learning_rate": 4.203819696267485e-09,
+      "loss": 0.5482,
+      "step": 593
+    },
+    {
+      "epoch": 1.9490968801313628,
+      "grad_norm": 17.418945183974255,
+      "learning_rate": 3.6950771532126003e-09,
+      "loss": 0.5732,
+      "step": 594
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 12.085404257651767,
+      "learning_rate": 3.219078458127078e-09,
+      "loss": 0.5876,
+      "step": 595
+    },
+    {
+      "epoch": 1.9556650246305418,
+      "grad_norm": 28.855096188397088,
+      "learning_rate": 2.775839254844925e-09,
+      "loss": 0.5689,
+      "step": 596
+    },
+    {
+      "epoch": 1.9589490968801315,
+      "grad_norm": 40.37266990923965,
+      "learning_rate": 2.3653741105499336e-09,
+      "loss": 0.5541,
+      "step": 597
+    },
+    {
+      "epoch": 1.9622331691297208,
+      "grad_norm": 55.51188079956238,
+      "learning_rate": 1.98769651529751e-09,
+      "loss": 0.5642,
+      "step": 598
+    },
+    {
+      "epoch": 1.9655172413793105,
+      "grad_norm": 34.1692976262017,
+      "learning_rate": 1.6428188815703626e-09,
+      "loss": 0.5595,
+      "step": 599
+    },
+    {
+      "epoch": 1.9688013136288998,
+      "grad_norm": 14.947354957753587,
+      "learning_rate": 1.330752543871161e-09,
+      "loss": 0.6005,
+      "step": 600
+    },
+    {
+      "epoch": 1.9720853858784895,
+      "grad_norm": 45.62738122621848,
+      "learning_rate": 1.0515077583498344e-09,
+      "loss": 0.5665,
+      "step": 601
+    },
+    {
+      "epoch": 1.9753694581280787,
+      "grad_norm": 21.725111315877975,
+      "learning_rate": 8.050937024666193e-10,
+      "loss": 0.5893,
+      "step": 602
+    },
+    {
+      "epoch": 1.9786535303776684,
+      "grad_norm": 35.375136168343055,
+      "learning_rate": 5.915184746904112e-10,
+      "loss": 0.5841,
+      "step": 603
+    },
+    {
+      "epoch": 1.9819376026272577,
+      "grad_norm": 18.604441345264426,
+      "learning_rate": 4.107890942325332e-10,
+      "loss": 0.5554,
+      "step": 604
+    },
+    {
+      "epoch": 1.9852216748768474,
+      "grad_norm": 119.72001100708987,
+      "learning_rate": 2.6291150081603207e-10,
+      "loss": 0.5614,
+      "step": 605
+    },
+    {
+      "epoch": 1.9885057471264367,
+      "grad_norm": 16.41601120115481,
+      "learning_rate": 1.4789055448061193e-10,
+      "loss": 0.5533,
+      "step": 606
+    },
+    {
+      "epoch": 1.9917898193760264,
+      "grad_norm": 124.04621529731233,
+      "learning_rate": 6.57300354227619e-11,
+      "loss": 0.5695,
+      "step": 607
+    },
+    {
+      "epoch": 1.9950738916256157,
+      "grad_norm": 41.47359974511061,
+      "learning_rate": 1.6432643871633346e-11,
+      "loss": 0.5409,
+      "step": 608
+    },
+    {
+      "epoch": 1.9950738916256157,
+      "eval_loss": 0.34100207686424255,
+      "eval_runtime": 255.0349,
+      "eval_samples_per_second": 8.101,
+      "eval_steps_per_second": 0.255,
+      "step": 608
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 608,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 152,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.488667559261635e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}