diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9086778736937755,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009086778736937755,
+      "grad_norm": 2.601068120514824,
+      "learning_rate": 0.0,
+      "loss": 0.7699,
+      "step": 1
+    },
+    {
+      "epoch": 0.001817355747387551,
+      "grad_norm": 2.545451150614842,
+      "learning_rate": 5.454545454545455e-07,
+      "loss": 0.7768,
+      "step": 2
+    },
+    {
+      "epoch": 0.0027260336210813267,
+      "grad_norm": 2.3373370819915653,
+      "learning_rate": 1.090909090909091e-06,
+      "loss": 0.7253,
+      "step": 3
+    },
+    {
+      "epoch": 0.003634711494775102,
+      "grad_norm": 2.48299255257347,
+      "learning_rate": 1.6363636363636363e-06,
+      "loss": 0.7588,
+      "step": 4
+    },
+    {
+      "epoch": 0.004543389368468878,
+      "grad_norm": 2.3266026791411267,
+      "learning_rate": 2.181818181818182e-06,
+      "loss": 0.7339,
+      "step": 5
+    },
+    {
+      "epoch": 0.005452067242162653,
+      "grad_norm": 2.3111442162120097,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.7585,
+      "step": 6
+    },
+    {
+      "epoch": 0.006360745115856429,
+      "grad_norm": 1.839056170375473,
+      "learning_rate": 3.2727272727272725e-06,
+      "loss": 0.7157,
+      "step": 7
+    },
+    {
+      "epoch": 0.007269422989550204,
+      "grad_norm": 1.8715018574536233,
+      "learning_rate": 3.818181818181818e-06,
+      "loss": 0.741,
+      "step": 8
+    },
+    {
+      "epoch": 0.00817810086324398,
+      "grad_norm": 1.6703798180686897,
+      "learning_rate": 4.363636363636364e-06,
+      "loss": 0.7339,
+      "step": 9
+    },
+    {
+      "epoch": 0.009086778736937756,
+      "grad_norm": 1.3022359540872304,
+      "learning_rate": 4.90909090909091e-06,
+      "loss": 0.7058,
+      "step": 10
+    },
+    {
+      "epoch": 0.009995456610631531,
+      "grad_norm": 1.1538753981218908,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.7158,
+      "step": 11
+    },
+    {
+      "epoch": 0.010904134484325307,
+      "grad_norm": 1.1727389823350258,
+      "learning_rate": 6e-06,
+      "loss": 0.7236,
+      "step": 12
+    },
+    {
+      "epoch": 0.011812812358019082,
+      "grad_norm": 1.7879982152520335,
+      "learning_rate": 6.545454545454545e-06,
+      "loss": 0.701,
+      "step": 13
+    },
+    {
+      "epoch": 0.012721490231712857,
+      "grad_norm": 1.8765006568093932,
+      "learning_rate": 7.090909090909091e-06,
+      "loss": 0.6837,
+      "step": 14
+    },
+    {
+      "epoch": 0.013630168105406633,
+      "grad_norm": 1.7017793683045053,
+      "learning_rate": 7.636363636363636e-06,
+      "loss": 0.6637,
+      "step": 15
+    },
+    {
+      "epoch": 0.014538845979100408,
+      "grad_norm": 1.5271998134211906,
+      "learning_rate": 8.181818181818181e-06,
+      "loss": 0.6793,
+      "step": 16
+    },
+    {
+      "epoch": 0.015447523852794184,
+      "grad_norm": 1.0560099784101875,
+      "learning_rate": 8.727272727272728e-06,
+      "loss": 0.6868,
+      "step": 17
+    },
+    {
+      "epoch": 0.01635620172648796,
+      "grad_norm": 1.1802463918516672,
+      "learning_rate": 9.272727272727273e-06,
+      "loss": 0.6477,
+      "step": 18
+    },
+    {
+      "epoch": 0.017264879600181735,
+      "grad_norm": 1.4964829988488886,
+      "learning_rate": 9.81818181818182e-06,
+      "loss": 0.6848,
+      "step": 19
+    },
+    {
+      "epoch": 0.01817355747387551,
+      "grad_norm": 1.0423576592539534,
+      "learning_rate": 1.0363636363636364e-05,
+      "loss": 0.6439,
+      "step": 20
+    },
+    {
+      "epoch": 0.019082235347569285,
+      "grad_norm": 0.8521617579985585,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.6197,
+      "step": 21
+    },
+    {
+      "epoch": 0.019990913221263062,
+      "grad_norm": 0.6455522796912817,
+      "learning_rate": 1.1454545454545455e-05,
+      "loss": 0.6165,
+      "step": 22
+    },
+    {
+      "epoch": 0.020899591094956836,
+      "grad_norm": 0.6424805984129104,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6447,
+      "step": 23
+    },
+    {
+      "epoch": 0.021808268968650613,
+      "grad_norm": 0.6452431266273964,
+      "learning_rate": 1.2545454545454545e-05,
+      "loss": 0.6248,
+      "step": 24
+    },
+    {
+      "epoch": 0.02271694684234439,
+      "grad_norm": 0.6091998561509787,
+      "learning_rate": 1.309090909090909e-05,
+      "loss": 0.6082,
+      "step": 25
+    },
+    {
+      "epoch": 0.023625624716038164,
+      "grad_norm": 0.5899107852705936,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.6381,
+      "step": 26
+    },
+    {
+      "epoch": 0.02453430258973194,
+      "grad_norm": 0.5221735021853063,
+      "learning_rate": 1.4181818181818181e-05,
+      "loss": 0.6375,
+      "step": 27
+    },
+    {
+      "epoch": 0.025442980463425715,
+      "grad_norm": 0.45451302628225215,
+      "learning_rate": 1.4727272727272728e-05,
+      "loss": 0.5926,
+      "step": 28
+    },
+    {
+      "epoch": 0.026351658337119492,
+      "grad_norm": 0.48697186829612116,
+      "learning_rate": 1.5272727272727273e-05,
+      "loss": 0.6094,
+      "step": 29
+    },
+    {
+      "epoch": 0.027260336210813266,
+      "grad_norm": 0.5191870380408595,
+      "learning_rate": 1.5818181818181818e-05,
+      "loss": 0.6339,
+      "step": 30
+    },
+    {
+      "epoch": 0.028169014084507043,
+      "grad_norm": 0.4558984969477796,
+      "learning_rate": 1.6363636363636363e-05,
+      "loss": 0.6045,
+      "step": 31
+    },
+    {
+      "epoch": 0.029077691958200817,
+      "grad_norm": 0.47078492141998163,
+      "learning_rate": 1.6909090909090907e-05,
+      "loss": 0.6168,
+      "step": 32
+    },
+    {
+      "epoch": 0.029986369831894594,
+      "grad_norm": 0.40599045996309796,
+      "learning_rate": 1.7454545454545456e-05,
+      "loss": 0.5991,
+      "step": 33
+    },
+    {
+      "epoch": 0.030895047705588367,
+      "grad_norm": 0.394694958480182,
+      "learning_rate": 1.8e-05,
+      "loss": 0.5896,
+      "step": 34
+    },
+    {
+      "epoch": 0.03180372557928214,
+      "grad_norm": 0.3940540280992479,
+      "learning_rate": 1.8545454545454545e-05,
+      "loss": 0.5873,
+      "step": 35
+    },
+    {
+      "epoch": 0.03271240345297592,
+      "grad_norm": 0.443078986005425,
+      "learning_rate": 1.909090909090909e-05,
+      "loss": 0.5893,
+      "step": 36
+    },
+    {
+      "epoch": 0.033621081326669695,
+      "grad_norm": 0.40061462050322844,
+      "learning_rate": 1.963636363636364e-05,
+      "loss": 0.6071,
+      "step": 37
+    },
+    {
+      "epoch": 0.03452975920036347,
+      "grad_norm": 0.3651344870337699,
+      "learning_rate": 2.0181818181818183e-05,
+      "loss": 0.5952,
+      "step": 38
+    },
+    {
+      "epoch": 0.03543843707405725,
+      "grad_norm": 0.3574739663311614,
+      "learning_rate": 2.0727272727272728e-05,
+      "loss": 0.5849,
+      "step": 39
+    },
+    {
+      "epoch": 0.03634711494775102,
+      "grad_norm": 0.39438420805485946,
+      "learning_rate": 2.1272727272727273e-05,
+      "loss": 0.5874,
+      "step": 40
+    },
+    {
+      "epoch": 0.0372557928214448,
+      "grad_norm": 0.38235263642566175,
+      "learning_rate": 2.1818181818181818e-05,
+      "loss": 0.589,
+      "step": 41
+    },
+    {
+      "epoch": 0.03816447069513857,
+      "grad_norm": 0.3166953970319303,
+      "learning_rate": 2.2363636363636366e-05,
+      "loss": 0.5442,
+      "step": 42
+    },
+    {
+      "epoch": 0.03907314856883235,
+      "grad_norm": 0.3831322680208187,
+      "learning_rate": 2.290909090909091e-05,
+      "loss": 0.5866,
+      "step": 43
+    },
+    {
+      "epoch": 0.039981826442526125,
+      "grad_norm": 0.40298176888334175,
+      "learning_rate": 2.3454545454545456e-05,
+      "loss": 0.5882,
+      "step": 44
+    },
+    {
+      "epoch": 0.0408905043162199,
+      "grad_norm": 0.3299018047565981,
+      "learning_rate": 2.4e-05,
+      "loss": 0.5837,
+      "step": 45
+    },
+    {
+      "epoch": 0.04179918218991367,
+      "grad_norm": 0.35755407226914043,
+      "learning_rate": 2.454545454545455e-05,
+      "loss": 0.6004,
+      "step": 46
+    },
+    {
+      "epoch": 0.04270786006360745,
+      "grad_norm": 0.40629650524300004,
+      "learning_rate": 2.509090909090909e-05,
+      "loss": 0.585,
+      "step": 47
+    },
+    {
+      "epoch": 0.04361653793730123,
+      "grad_norm": 0.35063710725122565,
+      "learning_rate": 2.5636363636363635e-05,
+      "loss": 0.556,
+      "step": 48
+    },
+    {
+      "epoch": 0.044525215810995,
+      "grad_norm": 0.34503154628666083,
+      "learning_rate": 2.618181818181818e-05,
+      "loss": 0.5638,
+      "step": 49
+    },
+    {
+      "epoch": 0.04543389368468878,
+      "grad_norm": 0.41060546364625117,
+      "learning_rate": 2.6727272727272728e-05,
+      "loss": 0.5799,
+      "step": 50
+    },
+    {
+      "epoch": 0.046342571558382555,
+      "grad_norm": 0.4194490541584686,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.5741,
+      "step": 51
+    },
+    {
+      "epoch": 0.04725124943207633,
+      "grad_norm": 0.3500886645834267,
+      "learning_rate": 2.7818181818181818e-05,
+      "loss": 0.5821,
+      "step": 52
+    },
+    {
+      "epoch": 0.0481599273057701,
+      "grad_norm": 0.3835932551976908,
+      "learning_rate": 2.8363636363636363e-05,
+      "loss": 0.5959,
+      "step": 53
+    },
+    {
+      "epoch": 0.04906860517946388,
+      "grad_norm": 0.3416728291114892,
+      "learning_rate": 2.890909090909091e-05,
+      "loss": 0.6064,
+      "step": 54
+    },
+    {
+      "epoch": 0.049977283053157656,
+      "grad_norm": 0.34404874346555325,
+      "learning_rate": 2.9454545454545456e-05,
+      "loss": 0.5755,
+      "step": 55
+    },
+    {
+      "epoch": 0.05088596092685143,
+      "grad_norm": 0.3710765456432596,
+      "learning_rate": 3e-05,
+      "loss": 0.5679,
+      "step": 56
+    },
+    {
+      "epoch": 0.051794638800545204,
+      "grad_norm": 0.29938282115592735,
+      "learning_rate": 2.9999932215858376e-05,
+      "loss": 0.5694,
+      "step": 57
+    },
+    {
+      "epoch": 0.052703316674238984,
+      "grad_norm": 0.35057253497259383,
+      "learning_rate": 2.9999728864046126e-05,
+      "loss": 0.5538,
+      "step": 58
+    },
+    {
+      "epoch": 0.05361199454793276,
+      "grad_norm": 0.30805671744258767,
+      "learning_rate": 2.9999389946401123e-05,
+      "loss": 0.5781,
+      "step": 59
+    },
+    {
+      "epoch": 0.05452067242162653,
+      "grad_norm": 0.3482706906480643,
+      "learning_rate": 2.9998915465986464e-05,
+      "loss": 0.5794,
+      "step": 60
+    },
+    {
+      "epoch": 0.05542935029532031,
+      "grad_norm": 0.337361409759639,
+      "learning_rate": 2.999830542709045e-05,
+      "loss": 0.5689,
+      "step": 61
+    },
+    {
+      "epoch": 0.056338028169014086,
+      "grad_norm": 0.3082843990483217,
+      "learning_rate": 2.9997559835226545e-05,
+      "loss": 0.5574,
+      "step": 62
+    },
+    {
+      "epoch": 0.05724670604270786,
+      "grad_norm": 0.3440993015680002,
+      "learning_rate": 2.9996678697133317e-05,
+      "loss": 0.5355,
+      "step": 63
+    },
+    {
+      "epoch": 0.05815538391640163,
+      "grad_norm": 0.3505726852328007,
+      "learning_rate": 2.9995662020774395e-05,
+      "loss": 0.5758,
+      "step": 64
+    },
+    {
+      "epoch": 0.059064061790095414,
+      "grad_norm": 0.3578683792155986,
+      "learning_rate": 2.999450981533838e-05,
+      "loss": 0.5576,
+      "step": 65
+    },
+    {
+      "epoch": 0.05997273966378919,
+      "grad_norm": 0.305534429017613,
+      "learning_rate": 2.999322209123878e-05,
+      "loss": 0.5611,
+      "step": 66
+    },
+    {
+      "epoch": 0.06088141753748296,
+      "grad_norm": 0.3461569434390687,
+      "learning_rate": 2.999179886011389e-05,
+      "loss": 0.5681,
+      "step": 67
+    },
+    {
+      "epoch": 0.061790095411176735,
+      "grad_norm": 0.36951699442069197,
+      "learning_rate": 2.999024013482672e-05,
+      "loss": 0.5546,
+      "step": 68
+    },
+    {
+      "epoch": 0.06269877328487052,
+      "grad_norm": 0.30035631327373347,
+      "learning_rate": 2.9988545929464837e-05,
+      "loss": 0.5641,
+      "step": 69
+    },
+    {
+      "epoch": 0.06360745115856428,
+      "grad_norm": 0.3147469851517606,
+      "learning_rate": 2.9986716259340288e-05,
+      "loss": 0.564,
+      "step": 70
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 0.3287019882975828,
+      "learning_rate": 2.9984751140989417e-05,
+      "loss": 0.5585,
+      "step": 71
+    },
+    {
+      "epoch": 0.06542480690595184,
+      "grad_norm": 0.3262923611974805,
+      "learning_rate": 2.9982650592172738e-05,
+      "loss": 0.5788,
+      "step": 72
+    },
+    {
+      "epoch": 0.06633348477964561,
+      "grad_norm": 0.30113514428264637,
+      "learning_rate": 2.998041463187477e-05,
+      "loss": 0.5397,
+      "step": 73
+    },
+    {
+      "epoch": 0.06724216265333939,
+      "grad_norm": 0.34137667654270437,
+      "learning_rate": 2.9978043280303867e-05,
+      "loss": 0.5695,
+      "step": 74
+    },
+    {
+      "epoch": 0.06815084052703317,
+      "grad_norm": 0.30136381991899835,
+      "learning_rate": 2.9975536558892034e-05,
+      "loss": 0.5574,
+      "step": 75
+    },
+    {
+      "epoch": 0.06905951840072694,
+      "grad_norm": 0.3029251079889505,
+      "learning_rate": 2.9972894490294738e-05,
+      "loss": 0.5519,
+      "step": 76
+    },
+    {
+      "epoch": 0.06996819627442072,
+      "grad_norm": 0.34180753067947,
+      "learning_rate": 2.9970117098390682e-05,
+      "loss": 0.5661,
+      "step": 77
+    },
+    {
+      "epoch": 0.0708768741481145,
+      "grad_norm": 0.3679070731789738,
+      "learning_rate": 2.9967204408281618e-05,
+      "loss": 0.5775,
+      "step": 78
+    },
+    {
+      "epoch": 0.07178555202180827,
+      "grad_norm": 0.29755135710898023,
+      "learning_rate": 2.9964156446292112e-05,
+      "loss": 0.5512,
+      "step": 79
+    },
+    {
+      "epoch": 0.07269422989550205,
+      "grad_norm": 0.3237156945925559,
+      "learning_rate": 2.9960973239969295e-05,
+      "loss": 0.5807,
+      "step": 80
+    },
+    {
+      "epoch": 0.07360290776919581,
+      "grad_norm": 0.28459486097868475,
+      "learning_rate": 2.9957654818082615e-05,
+      "loss": 0.5397,
+      "step": 81
+    },
+    {
+      "epoch": 0.0745115856428896,
+      "grad_norm": 0.3299635371208699,
+      "learning_rate": 2.9954201210623594e-05,
+      "loss": 0.5742,
+      "step": 82
+    },
+    {
+      "epoch": 0.07542026351658337,
+      "grad_norm": 0.29790933007572773,
+      "learning_rate": 2.995061244880554e-05,
+      "loss": 0.5498,
+      "step": 83
+    },
+    {
+      "epoch": 0.07632894139027714,
+      "grad_norm": 0.310578411796522,
+      "learning_rate": 2.994688856506327e-05,
+      "loss": 0.5772,
+      "step": 84
+    },
+    {
+      "epoch": 0.07723761926397092,
+      "grad_norm": 0.2830093354841121,
+      "learning_rate": 2.9943029593052822e-05,
+      "loss": 0.5612,
+      "step": 85
+    },
+    {
+      "epoch": 0.0781462971376647,
+      "grad_norm": 0.31044824437431984,
+      "learning_rate": 2.9939035567651146e-05,
+      "loss": 0.5494,
+      "step": 86
+    },
+    {
+      "epoch": 0.07905497501135847,
+      "grad_norm": 0.26786587945077706,
+      "learning_rate": 2.9934906524955777e-05,
+      "loss": 0.5115,
+      "step": 87
+    },
+    {
+      "epoch": 0.07996365288505225,
+      "grad_norm": 0.3128021372443758,
+      "learning_rate": 2.9930642502284537e-05,
+      "loss": 0.5531,
+      "step": 88
+    },
+    {
+      "epoch": 0.08087233075874603,
+      "grad_norm": 0.2775957207658622,
+      "learning_rate": 2.9926243538175172e-05,
+      "loss": 0.5387,
+      "step": 89
+    },
+    {
+      "epoch": 0.0817810086324398,
+      "grad_norm": 0.27919195499528654,
+      "learning_rate": 2.992170967238502e-05,
+      "loss": 0.5528,
+      "step": 90
+    },
+    {
+      "epoch": 0.08268968650613358,
+      "grad_norm": 0.3078667562238638,
+      "learning_rate": 2.9917040945890638e-05,
+      "loss": 0.545,
+      "step": 91
+    },
+    {
+      "epoch": 0.08359836437982734,
+      "grad_norm": 0.32246942789494776,
+      "learning_rate": 2.9912237400887442e-05,
+      "loss": 0.5741,
+      "step": 92
+    },
+    {
+      "epoch": 0.08450704225352113,
+      "grad_norm": 0.31869933415834445,
+      "learning_rate": 2.9907299080789326e-05,
+      "loss": 0.5655,
+      "step": 93
+    },
+    {
+      "epoch": 0.0854157201272149,
+      "grad_norm": 0.26749345526391427,
+      "learning_rate": 2.9902226030228252e-05,
+      "loss": 0.5255,
+      "step": 94
+    },
+    {
+      "epoch": 0.08632439800090867,
+      "grad_norm": 0.32883178074329494,
+      "learning_rate": 2.9897018295053883e-05,
+      "loss": 0.5249,
+      "step": 95
+    },
+    {
+      "epoch": 0.08723307587460245,
+      "grad_norm": 0.3232345235969363,
+      "learning_rate": 2.9891675922333125e-05,
+      "loss": 0.5692,
+      "step": 96
+    },
+    {
+      "epoch": 0.08814175374829623,
+      "grad_norm": 0.3548603927764629,
+      "learning_rate": 2.9886198960349733e-05,
+      "loss": 0.5688,
+      "step": 97
+    },
+    {
+      "epoch": 0.08905043162199,
+      "grad_norm": 0.4796806813236084,
+      "learning_rate": 2.9880587458603862e-05,
+      "loss": 0.5615,
+      "step": 98
+    },
+    {
+      "epoch": 0.08995910949568378,
+      "grad_norm": 0.3048044136414707,
+      "learning_rate": 2.9874841467811624e-05,
+      "loss": 0.5455,
+      "step": 99
+    },
+    {
+      "epoch": 0.09086778736937756,
+      "grad_norm": 0.3345696458929397,
+      "learning_rate": 2.9868961039904628e-05,
+      "loss": 0.5521,
+      "step": 100
+    },
+    {
+      "epoch": 0.09177646524307133,
+      "grad_norm": 0.29613055665568655,
+      "learning_rate": 2.9862946228029507e-05,
+      "loss": 0.5381,
+      "step": 101
+    },
+    {
+      "epoch": 0.09268514311676511,
+      "grad_norm": 0.3088460179022418,
+      "learning_rate": 2.9856797086547435e-05,
+      "loss": 0.5585,
+      "step": 102
+    },
+    {
+      "epoch": 0.09359382099045888,
+      "grad_norm": 0.34901311274948843,
+      "learning_rate": 2.9850513671033664e-05,
+      "loss": 0.573,
+      "step": 103
+    },
+    {
+      "epoch": 0.09450249886415266,
+      "grad_norm": 0.352416635014394,
+      "learning_rate": 2.984409603827697e-05,
+      "loss": 0.5635,
+      "step": 104
+    },
+    {
+      "epoch": 0.09541117673784644,
+      "grad_norm": 0.3207125359081875,
+      "learning_rate": 2.983754424627919e-05,
+      "loss": 0.5215,
+      "step": 105
+    },
+    {
+      "epoch": 0.0963198546115402,
+      "grad_norm": 0.31749439598402857,
+      "learning_rate": 2.9830858354254672e-05,
+      "loss": 0.541,
+      "step": 106
+    },
+    {
+      "epoch": 0.09722853248523398,
+      "grad_norm": 0.31449476495281475,
+      "learning_rate": 2.9824038422629737e-05,
+      "loss": 0.5506,
+      "step": 107
+    },
+    {
+      "epoch": 0.09813721035892777,
+      "grad_norm": 0.44629956406034843,
+      "learning_rate": 2.9817084513042153e-05,
+      "loss": 0.549,
+      "step": 108
+    },
+    {
+      "epoch": 0.09904588823262153,
+      "grad_norm": 0.3066824935289338,
+      "learning_rate": 2.9809996688340552e-05,
+      "loss": 0.5445,
+      "step": 109
+    },
+    {
+      "epoch": 0.09995456610631531,
+      "grad_norm": 0.34715367523179214,
+      "learning_rate": 2.9802775012583884e-05,
+      "loss": 0.5596,
+      "step": 110
+    },
+    {
+      "epoch": 0.10086324398000909,
+      "grad_norm": 0.30888532714559946,
+      "learning_rate": 2.9795419551040836e-05,
+      "loss": 0.556,
+      "step": 111
+    },
+    {
+      "epoch": 0.10177192185370286,
+      "grad_norm": 0.3527630189050702,
+      "learning_rate": 2.978793037018922e-05,
+      "loss": 0.5548,
+      "step": 112
+    },
+    {
+      "epoch": 0.10268059972739664,
+      "grad_norm": 0.3161408926482892,
+      "learning_rate": 2.9780307537715396e-05,
+      "loss": 0.5626,
+      "step": 113
+    },
+    {
+      "epoch": 0.10358927760109041,
+      "grad_norm": 0.28890312139326835,
+      "learning_rate": 2.9772551122513652e-05,
+      "loss": 0.5507,
+      "step": 114
+    },
+    {
+      "epoch": 0.10449795547478419,
+      "grad_norm": 0.3229834707417106,
+      "learning_rate": 2.9764661194685583e-05,
+      "loss": 0.5564,
+      "step": 115
+    },
+    {
+      "epoch": 0.10540663334847797,
+      "grad_norm": 0.31937480406323543,
+      "learning_rate": 2.9756637825539453e-05,
+      "loss": 0.5416,
+      "step": 116
+    },
+    {
+      "epoch": 0.10631531122217174,
+      "grad_norm": 0.28663393401926945,
+      "learning_rate": 2.9748481087589552e-05,
+      "loss": 0.5583,
+      "step": 117
+    },
+    {
+      "epoch": 0.10722398909586552,
+      "grad_norm": 0.2974995696933711,
+      "learning_rate": 2.974019105455554e-05,
+      "loss": 0.5499,
+      "step": 118
+    },
+    {
+      "epoch": 0.1081326669695593,
+      "grad_norm": 0.29684178337584144,
+      "learning_rate": 2.97317678013618e-05,
+      "loss": 0.5609,
+      "step": 119
+    },
+    {
+      "epoch": 0.10904134484325306,
+      "grad_norm": 0.3262506812045903,
+      "learning_rate": 2.972321140413672e-05,
+      "loss": 0.5455,
+      "step": 120
+    },
+    {
+      "epoch": 0.10995002271694684,
+      "grad_norm": 0.3092601543840332,
+      "learning_rate": 2.971452194021204e-05,
+      "loss": 0.551,
+      "step": 121
+    },
+    {
+      "epoch": 0.11085870059064062,
+      "grad_norm": 0.4005513894771114,
+      "learning_rate": 2.970569948812214e-05,
+      "loss": 0.5591,
+      "step": 122
+    },
+    {
+      "epoch": 0.11176737846433439,
+      "grad_norm": 0.327190640093946,
+      "learning_rate": 2.969674412760334e-05,
+      "loss": 0.5428,
+      "step": 123
+    },
+    {
+      "epoch": 0.11267605633802817,
+      "grad_norm": 0.3155091476237132,
+      "learning_rate": 2.968765593959315e-05,
+      "loss": 0.5497,
+      "step": 124
+    },
+    {
+      "epoch": 0.11358473421172194,
+      "grad_norm": 0.3624314389132723,
+      "learning_rate": 2.9678435006229585e-05,
+      "loss": 0.5599,
+      "step": 125
+    },
+    {
+      "epoch": 0.11449341208541572,
+      "grad_norm": 0.2696492296407551,
+      "learning_rate": 2.9669081410850378e-05,
+      "loss": 0.5466,
+      "step": 126
+    },
+    {
+      "epoch": 0.1154020899591095,
+      "grad_norm": 0.35239441892045026,
+      "learning_rate": 2.9659595237992256e-05,
+      "loss": 0.5677,
+      "step": 127
+    },
+    {
+      "epoch": 0.11631076783280327,
+      "grad_norm": 0.2770977804416505,
+      "learning_rate": 2.9649976573390168e-05,
+      "loss": 0.5551,
+      "step": 128
+    },
+    {
+      "epoch": 0.11721944570649705,
+      "grad_norm": 0.317546939102512,
+      "learning_rate": 2.9640225503976495e-05,
+      "loss": 0.5493,
+      "step": 129
+    },
+    {
+      "epoch": 0.11812812358019083,
+      "grad_norm": 0.29901361868733123,
+      "learning_rate": 2.9630342117880293e-05,
+      "loss": 0.5512,
+      "step": 130
+    },
+    {
+      "epoch": 0.1190368014538846,
+      "grad_norm": 0.34796394305794476,
+      "learning_rate": 2.9620326504426476e-05,
+      "loss": 0.5449,
+      "step": 131
+    },
+    {
+      "epoch": 0.11994547932757837,
+      "grad_norm": 0.3189637752329797,
+      "learning_rate": 2.9610178754135005e-05,
+      "loss": 0.5352,
+      "step": 132
+    },
+    {
+      "epoch": 0.12085415720127216,
+      "grad_norm": 0.3621323294601048,
+      "learning_rate": 2.9599898958720088e-05,
+      "loss": 0.5758,
+      "step": 133
+    },
+    {
+      "epoch": 0.12176283507496592,
+      "grad_norm": 0.3017668854168288,
+      "learning_rate": 2.958948721108934e-05,
+      "loss": 0.5524,
+      "step": 134
+    },
+    {
+      "epoch": 0.1226715129486597,
+      "grad_norm": 0.3019819675954715,
+      "learning_rate": 2.957894360534295e-05,
+      "loss": 0.5408,
+      "step": 135
+    },
+    {
+      "epoch": 0.12358019082235347,
+      "grad_norm": 0.29500722376897576,
+      "learning_rate": 2.9568268236772816e-05,
+      "loss": 0.5461,
+      "step": 136
+    },
+    {
+      "epoch": 0.12448886869604725,
+      "grad_norm": 0.4193960921261131,
+      "learning_rate": 2.955746120186169e-05,
+      "loss": 0.5355,
+      "step": 137
+    },
+    {
+      "epoch": 0.12539754656974103,
+      "grad_norm": 0.2750383897904465,
+      "learning_rate": 2.9546522598282325e-05,
+      "loss": 0.5596,
+      "step": 138
+    },
+    {
+      "epoch": 0.1263062244434348,
+      "grad_norm": 0.34998374547584626,
+      "learning_rate": 2.953545252489657e-05,
+      "loss": 0.5489,
+      "step": 139
+    },
+    {
+      "epoch": 0.12721490231712856,
+      "grad_norm": 0.2701830779505992,
+      "learning_rate": 2.9524251081754475e-05,
+      "loss": 0.5412,
+      "step": 140
+    },
+    {
+      "epoch": 0.12812358019082234,
+      "grad_norm": 0.29400344926785826,
+      "learning_rate": 2.9512918370093407e-05,
+      "loss": 0.5541,
+      "step": 141
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 0.2684006986647009,
+      "learning_rate": 2.9501454492337107e-05,
+      "loss": 0.5249,
+      "step": 142
+    },
+    {
+      "epoch": 0.1299409359382099,
+      "grad_norm": 0.34974547659671557,
+      "learning_rate": 2.9489859552094806e-05,
+      "loss": 0.5722,
+      "step": 143
+    },
+    {
+      "epoch": 0.1308496138119037,
+      "grad_norm": 0.2911364725934339,
+      "learning_rate": 2.947813365416023e-05,
+      "loss": 0.534,
+      "step": 144
+    },
+    {
+      "epoch": 0.13175829168559747,
+      "grad_norm": 0.2685521574942707,
+      "learning_rate": 2.9466276904510713e-05,
+      "loss": 0.5282,
+      "step": 145
+    },
+    {
+      "epoch": 0.13266696955929122,
+      "grad_norm": 0.27716261776006357,
+      "learning_rate": 2.9454289410306202e-05,
+      "loss": 0.5559,
+      "step": 146
+    },
+    {
+      "epoch": 0.133575647432985,
+      "grad_norm": 0.32403029678097417,
+      "learning_rate": 2.9442171279888286e-05,
+      "loss": 0.5562,
+      "step": 147
+    },
+    {
+      "epoch": 0.13448432530667878,
+      "grad_norm": 0.29914381494298525,
+      "learning_rate": 2.942992262277926e-05,
+      "loss": 0.5498,
+      "step": 148
+    },
+    {
+      "epoch": 0.13539300318037256,
+      "grad_norm": 0.2999453587781109,
+      "learning_rate": 2.9417543549681067e-05,
+      "loss": 0.5512,
+      "step": 149
+    },
+    {
+      "epoch": 0.13630168105406634,
+      "grad_norm": 0.33912555060096045,
+      "learning_rate": 2.9405034172474363e-05,
+      "loss": 0.5513,
+      "step": 150
+    },
+    {
+      "epoch": 0.1372103589277601,
+      "grad_norm": 0.2786003018351176,
+      "learning_rate": 2.939239460421746e-05,
+      "loss": 0.5357,
+      "step": 151
+    },
+    {
+      "epoch": 0.13811903680145388,
+      "grad_norm": 0.31469991721983587,
+      "learning_rate": 2.937962495914534e-05,
+      "loss": 0.5471,
+      "step": 152
+    },
+    {
+      "epoch": 0.13902771467514766,
+      "grad_norm": 0.3238481784334065,
+      "learning_rate": 2.9366725352668584e-05,
+      "loss": 0.5345,
+      "step": 153
+    },
+    {
+      "epoch": 0.13993639254884144,
+      "grad_norm": 0.28913738590561583,
+      "learning_rate": 2.9353695901372363e-05,
+      "loss": 0.5413,
+      "step": 154
+    },
+    {
+      "epoch": 0.14084507042253522,
+      "grad_norm": 0.2985726097232229,
+      "learning_rate": 2.9340536723015367e-05,
+      "loss": 0.5422,
+      "step": 155
+    },
+    {
+      "epoch": 0.141753748296229,
+      "grad_norm": 0.3520712671470228,
+      "learning_rate": 2.9327247936528742e-05,
+      "loss": 0.5137,
+      "step": 156
+    },
+    {
+      "epoch": 0.14266242616992275,
+      "grad_norm": 0.27128214863020367,
+      "learning_rate": 2.931382966201502e-05,
+      "loss": 0.5434,
+      "step": 157
+    },
+    {
+      "epoch": 0.14357110404361653,
+      "grad_norm": 0.29607019876927254,
+      "learning_rate": 2.930028202074703e-05,
+      "loss": 0.5333,
+      "step": 158
+    },
+    {
+      "epoch": 0.1444797819173103,
+      "grad_norm": 0.3443171478381605,
+      "learning_rate": 2.9286605135166806e-05,
+      "loss": 0.5429,
+      "step": 159
+    },
+    {
+      "epoch": 0.1453884597910041,
+      "grad_norm": 0.2724745024955907,
+      "learning_rate": 2.927279912888447e-05,
+      "loss": 0.5436,
+      "step": 160
+    },
+    {
+      "epoch": 0.14629713766469787,
+      "grad_norm": 0.32926171965555123,
+      "learning_rate": 2.9258864126677132e-05,
+      "loss": 0.5457,
+      "step": 161
+    },
+    {
+      "epoch": 0.14720581553839163,
+      "grad_norm": 0.26119164354842045,
+      "learning_rate": 2.9244800254487744e-05,
+      "loss": 0.5366,
+      "step": 162
+    },
+    {
+      "epoch": 0.1481144934120854,
+      "grad_norm": 0.3048750099033153,
+      "learning_rate": 2.9230607639423973e-05,
+      "loss": 0.539,
+      "step": 163
+    },
+    {
+      "epoch": 0.1490231712857792,
+      "grad_norm": 0.3235214935466523,
+      "learning_rate": 2.9216286409757052e-05,
+      "loss": 0.5447,
+      "step": 164
+    },
+    {
+      "epoch": 0.14993184915947297,
+      "grad_norm": 0.37561127087178964,
+      "learning_rate": 2.920183669492061e-05,
+      "loss": 0.5611,
+      "step": 165
+    },
+    {
+      "epoch": 0.15084052703316675,
+      "grad_norm": 0.300540850326533,
+      "learning_rate": 2.9187258625509518e-05,
+      "loss": 0.5367,
+      "step": 166
+    },
+    {
+      "epoch": 0.15174920490686053,
+      "grad_norm": 0.3576135446433189,
+      "learning_rate": 2.9172552333278708e-05,
+      "loss": 0.5337,
+      "step": 167
+    },
+    {
+      "epoch": 0.15265788278055428,
+      "grad_norm": 0.31274340375293364,
+      "learning_rate": 2.9157717951141953e-05,
+      "loss": 0.5381,
+      "step": 168
+    },
+    {
+      "epoch": 0.15356656065424806,
+      "grad_norm": 0.3352810642235237,
+      "learning_rate": 2.914275561317071e-05,
+      "loss": 0.548,
+      "step": 169
+    },
+    {
+      "epoch": 0.15447523852794184,
+      "grad_norm": 0.30963404928391625,
+      "learning_rate": 2.9127665454592872e-05,
+      "loss": 0.5345,
+      "step": 170
+    },
+    {
+      "epoch": 0.15538391640163562,
+      "grad_norm": 0.3071782564022272,
+      "learning_rate": 2.9112447611791563e-05,
+      "loss": 0.5298,
+      "step": 171
+    },
+    {
+      "epoch": 0.1562925942753294,
+      "grad_norm": 0.29846884519248473,
+      "learning_rate": 2.9097102222303914e-05,
+      "loss": 0.5373,
+      "step": 172
+    },
+    {
+      "epoch": 0.15720127214902316,
+      "grad_norm": 0.32363500369528425,
+      "learning_rate": 2.9081629424819792e-05,
+      "loss": 0.5419,
+      "step": 173
+    },
+    {
+      "epoch": 0.15810995002271694,
+      "grad_norm": 0.2979588134888228,
+      "learning_rate": 2.906602935918057e-05,
+      "loss": 0.5481,
+      "step": 174
+    },
+    {
+      "epoch": 0.15901862789641072,
+      "grad_norm": 0.30814351362966413,
+      "learning_rate": 2.9050302166377858e-05,
+      "loss": 0.5226,
+      "step": 175
+    },
+    {
+      "epoch": 0.1599273057701045,
+      "grad_norm": 0.2842362351308814,
+      "learning_rate": 2.9034447988552227e-05,
+      "loss": 0.5466,
+      "step": 176
+    },
+    {
+      "epoch": 0.16083598364379828,
+      "grad_norm": 0.30495693391974155,
+      "learning_rate": 2.9018466968991913e-05,
+      "loss": 0.5293,
+      "step": 177
+    },
+    {
+      "epoch": 0.16174466151749206,
+      "grad_norm": 0.2972660949512962,
+      "learning_rate": 2.9002359252131547e-05,
+      "loss": 0.5373,
+      "step": 178
+    },
+    {
+      "epoch": 0.1626533393911858,
+      "grad_norm": 0.25956110689877127,
+      "learning_rate": 2.8986124983550836e-05,
+      "loss": 0.5409,
+      "step": 179
+    },
+    {
+      "epoch": 0.1635620172648796,
+      "grad_norm": 0.28704003352350194,
+      "learning_rate": 2.896976430997323e-05,
+      "loss": 0.5367,
+      "step": 180
+    },
+    {
+      "epoch": 0.16447069513857338,
+      "grad_norm": 0.27494590106809935,
+      "learning_rate": 2.8953277379264633e-05,
+      "loss": 0.5245,
+      "step": 181
+    },
+    {
+      "epoch": 0.16537937301226716,
+      "grad_norm": 0.29025723129556125,
+      "learning_rate": 2.8936664340432033e-05,
+      "loss": 0.5634,
+      "step": 182
+    },
+    {
+      "epoch": 0.16628805088596094,
+      "grad_norm": 0.2640721729777581,
+      "learning_rate": 2.891992534362218e-05,
+      "loss": 0.5575,
+      "step": 183
+    },
+    {
+      "epoch": 0.1671967287596547,
+      "grad_norm": 0.28770467313834736,
+      "learning_rate": 2.8903060540120203e-05,
+      "loss": 0.5526,
+      "step": 184
+    },
+    {
+      "epoch": 0.16810540663334847,
+      "grad_norm": 0.267891776899334,
+      "learning_rate": 2.8886070082348268e-05,
+      "loss": 0.5475,
+      "step": 185
+    },
+    {
+      "epoch": 0.16901408450704225,
+      "grad_norm": 0.3035468603633203,
+      "learning_rate": 2.8868954123864194e-05,
+      "loss": 0.5288,
+      "step": 186
+    },
+    {
+      "epoch": 0.16992276238073603,
+      "grad_norm": 0.2597479657341249,
+      "learning_rate": 2.885171281936005e-05,
+      "loss": 0.5221,
+      "step": 187
+    },
+    {
+      "epoch": 0.1708314402544298,
+      "grad_norm": 0.30549990341073835,
+      "learning_rate": 2.883434632466077e-05,
+      "loss": 0.5241,
+      "step": 188
+    },
+    {
+      "epoch": 0.1717401181281236,
+      "grad_norm": 0.3097670337344422,
+      "learning_rate": 2.8816854796722754e-05,
+      "loss": 0.5567,
+      "step": 189
+    },
+    {
+      "epoch": 0.17264879600181735,
+      "grad_norm": 0.24887750502737235,
+      "learning_rate": 2.879923839363242e-05,
+      "loss": 0.554,
+      "step": 190
+    },
+    {
+      "epoch": 0.17355747387551113,
+      "grad_norm": 0.30369106518383837,
+      "learning_rate": 2.878149727460481e-05,
+      "loss": 0.5322,
+      "step": 191
+    },
+    {
+      "epoch": 0.1744661517492049,
+      "grad_norm": 0.2734596089931484,
+      "learning_rate": 2.8763631599982126e-05,
+      "loss": 0.5345,
+      "step": 192
+    },
+    {
+      "epoch": 0.1753748296228987,
+      "grad_norm": 0.29225021074746216,
+      "learning_rate": 2.874564153123228e-05,
+      "loss": 0.5419,
+      "step": 193
+    },
+    {
+      "epoch": 0.17628350749659247,
+      "grad_norm": 0.3002278845544173,
+      "learning_rate": 2.8727527230947473e-05,
+      "loss": 0.5326,
+      "step": 194
+    },
+    {
+      "epoch": 0.17719218537028622,
+      "grad_norm": 0.2794201841833264,
+      "learning_rate": 2.870928886284267e-05,
+      "loss": 0.5251,
+      "step": 195
+    },
+    {
+      "epoch": 0.17810086324398,
+      "grad_norm": 0.2811482022282061,
+      "learning_rate": 2.8690926591754142e-05,
+      "loss": 0.5485,
+      "step": 196
+    },
+    {
+      "epoch": 0.17900954111767378,
+      "grad_norm": 0.2911471546431618,
+      "learning_rate": 2.8672440583638013e-05,
+      "loss": 0.5407,
+      "step": 197
+    },
+    {
+      "epoch": 0.17991821899136756,
+      "grad_norm": 0.2730896343056817,
+      "learning_rate": 2.86538310055687e-05,
+      "loss": 0.5047,
+      "step": 198
+    },
+    {
+      "epoch": 0.18082689686506134,
+      "grad_norm": 0.3138295451241485,
+      "learning_rate": 2.863509802573744e-05,
+      "loss": 0.558,
+      "step": 199
+    },
+    {
+      "epoch": 0.18173557473875512,
+      "grad_norm": 0.30518868956972217,
+      "learning_rate": 2.8616241813450755e-05,
+      "loss": 0.533,
+      "step": 200
+    },
+    {
+      "epoch": 0.18264425261244888,
+      "grad_norm": 0.2746559666747863,
+      "learning_rate": 2.8597262539128947e-05,
+      "loss": 0.5207,
+      "step": 201
+    },
+    {
+      "epoch": 0.18355293048614266,
+      "grad_norm": 0.2982492469468351,
+      "learning_rate": 2.857816037430451e-05,
+      "loss": 0.5309,
+      "step": 202
+    },
+    {
+      "epoch": 0.18446160835983644,
+      "grad_norm": 0.2819015087920199,
+      "learning_rate": 2.8558935491620634e-05,
+      "loss": 0.5414,
+      "step": 203
+    },
+    {
+      "epoch": 0.18537028623353022,
+      "grad_norm": 0.2831047957568758,
+      "learning_rate": 2.8539588064829605e-05,
+      "loss": 0.5539,
+      "step": 204
+    },
+    {
+      "epoch": 0.186278964107224,
+      "grad_norm": 0.3451150482089612,
+      "learning_rate": 2.852011826879125e-05,
+      "loss": 0.5331,
+      "step": 205
+    },
+    {
+      "epoch": 0.18718764198091775,
+      "grad_norm": 0.2525444743911961,
+      "learning_rate": 2.8500526279471362e-05,
+      "loss": 0.5272,
+      "step": 206
+    },
+    {
+      "epoch": 0.18809631985461153,
+      "grad_norm": 0.29713374816520677,
+      "learning_rate": 2.8480812273940097e-05,
+      "loss": 0.5526,
+      "step": 207
+    },
+    {
+      "epoch": 0.1890049977283053,
+      "grad_norm": 0.2685082552759803,
+      "learning_rate": 2.8460976430370375e-05,
+      "loss": 0.5114,
+      "step": 208
+    },
+    {
+      "epoch": 0.1899136756019991,
+      "grad_norm": 0.2784798341003174,
+      "learning_rate": 2.8441018928036287e-05,
+      "loss": 0.5204,
+      "step": 209
+    },
+    {
+      "epoch": 0.19082235347569287,
+      "grad_norm": 0.30091412799694445,
+      "learning_rate": 2.8420939947311454e-05,
+      "loss": 0.5371,
+      "step": 210
+    },
+    {
+      "epoch": 0.19173103134938665,
+      "grad_norm": 0.28284608142963674,
+      "learning_rate": 2.84007396696674e-05,
+      "loss": 0.5274,
+      "step": 211
+    },
+    {
+      "epoch": 0.1926397092230804,
+      "grad_norm": 0.3282616006754972,
+      "learning_rate": 2.8380418277671928e-05,
+      "loss": 0.5319,
+      "step": 212
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 0.26486949289774536,
+      "learning_rate": 2.8359975954987453e-05,
+      "loss": 0.5319,
+      "step": 213
+    },
+    {
+      "epoch": 0.19445706497046797,
+      "grad_norm": 0.3161233971442626,
+      "learning_rate": 2.8339412886369345e-05,
+      "loss": 0.5362,
+      "step": 214
+    },
+    {
+      "epoch": 0.19536574284416175,
+      "grad_norm": 0.299905638785644,
+      "learning_rate": 2.8318729257664265e-05,
+      "loss": 0.5632,
+      "step": 215
+    },
+    {
+      "epoch": 0.19627442071785553,
+      "grad_norm": 0.27636515077908513,
+      "learning_rate": 2.8297925255808484e-05,
+      "loss": 0.5359,
+      "step": 216
+    },
+    {
+      "epoch": 0.19718309859154928,
+      "grad_norm": 0.33338932554796485,
+      "learning_rate": 2.8277001068826187e-05,
+      "loss": 0.5421,
+      "step": 217
+    },
+    {
+      "epoch": 0.19809177646524306,
+      "grad_norm": 0.27956404676663693,
+      "learning_rate": 2.8255956885827786e-05,
+      "loss": 0.5309,
+      "step": 218
+    },
+    {
+      "epoch": 0.19900045433893684,
+      "grad_norm": 0.28929701016126264,
+      "learning_rate": 2.8234792897008194e-05,
+      "loss": 0.5482,
+      "step": 219
+    },
+    {
+      "epoch": 0.19990913221263062,
+      "grad_norm": 0.2937742621819467,
+      "learning_rate": 2.821350929364512e-05,
+      "loss": 0.537,
+      "step": 220
+    },
+    {
+      "epoch": 0.2008178100863244,
+      "grad_norm": 0.27127495025152987,
+      "learning_rate": 2.8192106268097336e-05,
+      "loss": 0.5271,
+      "step": 221
+    },
+    {
+      "epoch": 0.20172648796001819,
+      "grad_norm": 0.26418445759257114,
+      "learning_rate": 2.817058401380294e-05,
+      "loss": 0.5251,
+      "step": 222
+    },
+    {
+      "epoch": 0.20263516583371194,
+      "grad_norm": 0.2858673177684851,
+      "learning_rate": 2.81489427252776e-05,
+      "loss": 0.5339,
+      "step": 223
+    },
+    {
+      "epoch": 0.20354384370740572,
+      "grad_norm": 0.25280337401887915,
+      "learning_rate": 2.8127182598112805e-05,
+      "loss": 0.5293,
+      "step": 224
+    },
+    {
+      "epoch": 0.2044525215810995,
+      "grad_norm": 0.29393678976163556,
+      "learning_rate": 2.8105303828974095e-05,
+      "loss": 0.5205,
+      "step": 225
+    },
+    {
+      "epoch": 0.20536119945479328,
+      "grad_norm": 0.27668804080319737,
+      "learning_rate": 2.8083306615599283e-05,
+      "loss": 0.5596,
+      "step": 226
+    },
+    {
+      "epoch": 0.20626987732848706,
+      "grad_norm": 0.2579574523472542,
+      "learning_rate": 2.8061191156796658e-05,
+      "loss": 0.5292,
+      "step": 227
+    },
+    {
+      "epoch": 0.20717855520218081,
+      "grad_norm": 0.2965602985642256,
+      "learning_rate": 2.8038957652443217e-05,
+      "loss": 0.5408,
+      "step": 228
+    },
+    {
+      "epoch": 0.2080872330758746,
+      "grad_norm": 0.2584149229749058,
+      "learning_rate": 2.801660630348282e-05,
+      "loss": 0.5219,
+      "step": 229
+    },
+    {
+      "epoch": 0.20899591094956838,
+      "grad_norm": 0.23617123717584307,
+      "learning_rate": 2.79941373119244e-05,
+      "loss": 0.5091,
+      "step": 230
+    },
+    {
+      "epoch": 0.20990458882326216,
+      "grad_norm": 0.29559409833291433,
+      "learning_rate": 2.7971550880840138e-05,
+      "loss": 0.5252,
+      "step": 231
+    },
+    {
+      "epoch": 0.21081326669695594,
+      "grad_norm": 0.25262076589345694,
+      "learning_rate": 2.794884721436361e-05,
+      "loss": 0.5322,
+      "step": 232
+    },
+    {
+      "epoch": 0.21172194457064972,
+      "grad_norm": 0.2861284017299984,
+      "learning_rate": 2.792602651768795e-05,
+      "loss": 0.519,
+      "step": 233
+    },
+    {
+      "epoch": 0.21263062244434347,
+      "grad_norm": 0.2940791172624826,
+      "learning_rate": 2.790308899706401e-05,
+      "loss": 0.5319,
+      "step": 234
+    },
+    {
+      "epoch": 0.21353930031803725,
+      "grad_norm": 0.24143923336501463,
+      "learning_rate": 2.7880034859798476e-05,
+      "loss": 0.5161,
+      "step": 235
+    },
+    {
+      "epoch": 0.21444797819173103,
+      "grad_norm": 0.2995164890461248,
+      "learning_rate": 2.7856864314251994e-05,
+      "loss": 0.5168,
+      "step": 236
+    },
+    {
+      "epoch": 0.2153566560654248,
+      "grad_norm": 0.25742178571208674,
+      "learning_rate": 2.783357756983731e-05,
+      "loss": 0.5534,
+      "step": 237
+    },
+    {
+      "epoch": 0.2162653339391186,
+      "grad_norm": 0.29245573697234445,
+      "learning_rate": 2.781017483701735e-05,
+      "loss": 0.5388,
+      "step": 238
+    },
+    {
+      "epoch": 0.21717401181281235,
+      "grad_norm": 0.2674112998244982,
+      "learning_rate": 2.7786656327303342e-05,
+      "loss": 0.5605,
+      "step": 239
+    },
+    {
+      "epoch": 0.21808268968650613,
+      "grad_norm": 0.263962722421515,
+      "learning_rate": 2.7763022253252882e-05,
+      "loss": 0.5283,
+      "step": 240
+    },
+    {
+      "epoch": 0.2189913675601999,
+      "grad_norm": 0.2960952065569361,
+      "learning_rate": 2.7739272828468022e-05,
+      "loss": 0.5346,
+      "step": 241
+    },
+    {
+      "epoch": 0.2199000454338937,
+      "grad_norm": 0.2564971928759544,
+      "learning_rate": 2.771540826759335e-05,
+      "loss": 0.5258,
+      "step": 242
+    },
+    {
+      "epoch": 0.22080872330758747,
+      "grad_norm": 0.38221953093331257,
+      "learning_rate": 2.769142878631403e-05,
+      "loss": 0.5386,
+      "step": 243
+    },
+    {
+      "epoch": 0.22171740118128125,
+      "grad_norm": 0.2567622031991432,
+      "learning_rate": 2.766733460135388e-05,
+      "loss": 0.5267,
+      "step": 244
+    },
+    {
+      "epoch": 0.222626079054975,
+      "grad_norm": 0.2810737233839325,
+      "learning_rate": 2.7643125930473375e-05,
+      "loss": 0.5378,
+      "step": 245
+    },
+    {
+      "epoch": 0.22353475692866878,
+      "grad_norm": 0.24712524183510226,
+      "learning_rate": 2.7618802992467718e-05,
+      "loss": 0.5129,
+      "step": 246
+    },
+    {
+      "epoch": 0.22444343480236256,
+      "grad_norm": 0.2581811828213249,
+      "learning_rate": 2.759436600716484e-05,
+      "loss": 0.5475,
+      "step": 247
+    },
+    {
+      "epoch": 0.22535211267605634,
+      "grad_norm": 0.26523091241064817,
+      "learning_rate": 2.756981519542341e-05,
+      "loss": 0.5161,
+      "step": 248
+    },
+    {
+      "epoch": 0.22626079054975012,
+      "grad_norm": 0.2747890186827313,
+      "learning_rate": 2.7545150779130858e-05,
+      "loss": 0.557,
+      "step": 249
+    },
+    {
+      "epoch": 0.22716946842344388,
+      "grad_norm": 0.28216298108132126,
+      "learning_rate": 2.7520372981201362e-05,
+      "loss": 0.529,
+      "step": 250
+    },
+    {
+      "epoch": 0.22807814629713766,
+      "grad_norm": 0.2827987705048485,
+      "learning_rate": 2.7495482025573817e-05,
+      "loss": 0.5326,
+      "step": 251
+    },
+    {
+      "epoch": 0.22898682417083144,
+      "grad_norm": 0.2588695610411959,
+      "learning_rate": 2.747047813720983e-05,
+      "loss": 0.5248,
+      "step": 252
+    },
+    {
+      "epoch": 0.22989550204452522,
+      "grad_norm": 0.27906004523644146,
+      "learning_rate": 2.7445361542091687e-05,
+      "loss": 0.5424,
+      "step": 253
+    },
+    {
+      "epoch": 0.230804179918219,
+      "grad_norm": 0.25979694399888736,
+      "learning_rate": 2.74201324672203e-05,
+      "loss": 0.5427,
+      "step": 254
+    },
+    {
+      "epoch": 0.23171285779191278,
+      "grad_norm": 0.3338334113235304,
+      "learning_rate": 2.739479114061316e-05,
+      "loss": 0.5407,
+      "step": 255
+    },
+    {
+      "epoch": 0.23262153566560653,
+      "grad_norm": 0.3034660576427874,
+      "learning_rate": 2.7369337791302272e-05,
+      "loss": 0.5381,
+      "step": 256
+    },
+    {
+      "epoch": 0.2335302135393003,
+      "grad_norm": 0.2770064979315646,
+      "learning_rate": 2.7343772649332097e-05,
+      "loss": 0.5628,
+      "step": 257
+    },
+    {
+      "epoch": 0.2344388914129941,
+      "grad_norm": 0.3235843684101457,
+      "learning_rate": 2.7318095945757465e-05,
+      "loss": 0.5488,
+      "step": 258
+    },
+    {
+      "epoch": 0.23534756928668787,
+      "grad_norm": 0.2656277121687186,
+      "learning_rate": 2.729230791264148e-05,
+      "loss": 0.5386,
+      "step": 259
+    },
+    {
+      "epoch": 0.23625624716038165,
+      "grad_norm": 0.2629765514542018,
+      "learning_rate": 2.7266408783053437e-05,
+      "loss": 0.5301,
+      "step": 260
+    },
+    {
+      "epoch": 0.2371649250340754,
+      "grad_norm": 0.2977558536425956,
+      "learning_rate": 2.72403987910667e-05,
+      "loss": 0.5282,
+      "step": 261
+    },
+    {
+      "epoch": 0.2380736029077692,
+      "grad_norm": 0.26390918112160183,
+      "learning_rate": 2.7214278171756602e-05,
+      "loss": 0.5534,
+      "step": 262
+    },
+    {
+      "epoch": 0.23898228078146297,
+      "grad_norm": 0.2913114222615873,
+      "learning_rate": 2.7188047161198318e-05,
+      "loss": 0.5296,
+      "step": 263
+    },
+    {
+      "epoch": 0.23989095865515675,
+      "grad_norm": 0.2843782894394526,
+      "learning_rate": 2.716170599646471e-05,
+      "loss": 0.5352,
+      "step": 264
+    },
+    {
+      "epoch": 0.24079963652885053,
+      "grad_norm": 0.28784903348306884,
+      "learning_rate": 2.7135254915624213e-05,
+      "loss": 0.5135,
+      "step": 265
+    },
+    {
+      "epoch": 0.2417083144025443,
+      "grad_norm": 0.282373175191605,
+      "learning_rate": 2.710869415773867e-05,
+      "loss": 0.5368,
+      "step": 266
+    },
+    {
+      "epoch": 0.24261699227623806,
+      "grad_norm": 0.27538619157579186,
+      "learning_rate": 2.7082023962861164e-05,
+      "loss": 0.5285,
+      "step": 267
+    },
+    {
+      "epoch": 0.24352567014993184,
+      "grad_norm": 0.26568964109400045,
+      "learning_rate": 2.7055244572033877e-05,
+      "loss": 0.5343,
+      "step": 268
+    },
+    {
+      "epoch": 0.24443434802362562,
+      "grad_norm": 0.27304231694701525,
+      "learning_rate": 2.702835622728587e-05,
+      "loss": 0.5247,
+      "step": 269
+    },
+    {
+      "epoch": 0.2453430258973194,
+      "grad_norm": 0.2712736651714342,
+      "learning_rate": 2.700135917163092e-05,
+      "loss": 0.5438,
+      "step": 270
+    },
+    {
+      "epoch": 0.2462517037710132,
+      "grad_norm": 0.27996658794844825,
+      "learning_rate": 2.697425364906534e-05,
+      "loss": 0.5343,
+      "step": 271
+    },
+    {
+      "epoch": 0.24716038164470694,
+      "grad_norm": 0.2734492688955939,
+      "learning_rate": 2.694703990456573e-05,
+      "loss": 0.5215,
+      "step": 272
+    },
+    {
+      "epoch": 0.24806905951840072,
+      "grad_norm": 0.25090613623072067,
+      "learning_rate": 2.69197181840868e-05,
+      "loss": 0.5077,
+      "step": 273
+    },
+    {
+      "epoch": 0.2489777373920945,
+      "grad_norm": 0.26818173165244835,
+      "learning_rate": 2.689228873455914e-05,
+      "loss": 0.5563,
+      "step": 274
+    },
+    {
+      "epoch": 0.24988641526578828,
+      "grad_norm": 0.29301009152758006,
+      "learning_rate": 2.6864751803886963e-05,
+      "loss": 0.5164,
+      "step": 275
+    },
+    {
+      "epoch": 0.25079509313948206,
+      "grad_norm": 0.27801377651380216,
+      "learning_rate": 2.6837107640945904e-05,
+      "loss": 0.5234,
+      "step": 276
+    },
+    {
+      "epoch": 0.2517037710131758,
+      "grad_norm": 0.2757615453146622,
+      "learning_rate": 2.6809356495580747e-05,
+      "loss": 0.5019,
+      "step": 277
+    },
+    {
+      "epoch": 0.2526124488868696,
+      "grad_norm": 0.28705078185667204,
+      "learning_rate": 2.6781498618603163e-05,
+      "loss": 0.5428,
+      "step": 278
+    },
+    {
+      "epoch": 0.2535211267605634,
+      "grad_norm": 0.27743176091962085,
+      "learning_rate": 2.6753534261789454e-05,
+      "loss": 0.5282,
+      "step": 279
+    },
+    {
+      "epoch": 0.25442980463425713,
+      "grad_norm": 0.2781590591093593,
+      "learning_rate": 2.672546367787828e-05,
+      "loss": 0.533,
+      "step": 280
+    },
+    {
+      "epoch": 0.25533848250795094,
+      "grad_norm": 0.26269046529463025,
+      "learning_rate": 2.6697287120568364e-05,
+      "loss": 0.534,
+      "step": 281
+    },
+    {
+      "epoch": 0.2562471603816447,
+      "grad_norm": 0.2976310747283509,
+      "learning_rate": 2.666900484451621e-05,
+      "loss": 0.5485,
+      "step": 282
+    },
+    {
+      "epoch": 0.2571558382553385,
+      "grad_norm": 0.27008030574397085,
+      "learning_rate": 2.664061710533379e-05,
+      "loss": 0.5378,
+      "step": 283
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 0.31570883702046254,
+      "learning_rate": 2.661212415958624e-05,
+      "loss": 0.5487,
+      "step": 284
+    },
+    {
+      "epoch": 0.25897319400272606,
+      "grad_norm": 0.2792688424866495,
+      "learning_rate": 2.6583526264789538e-05,
+      "loss": 0.5446,
+      "step": 285
+    },
+    {
+      "epoch": 0.2598818718764198,
+      "grad_norm": 0.2777118844456695,
+      "learning_rate": 2.6554823679408195e-05,
+      "loss": 0.5023,
+      "step": 286
+    },
+    {
+      "epoch": 0.26079054975011356,
+      "grad_norm": 0.31599907246218734,
+      "learning_rate": 2.6526016662852887e-05,
+      "loss": 0.5239,
+      "step": 287
+    },
+    {
+      "epoch": 0.2616992276238074,
+      "grad_norm": 0.26231651760848274,
+      "learning_rate": 2.649710547547813e-05,
+      "loss": 0.5261,
+      "step": 288
+    },
+    {
+      "epoch": 0.2626079054975011,
+      "grad_norm": 0.3981071563722695,
+      "learning_rate": 2.6468090378579933e-05,
+      "loss": 0.5549,
+      "step": 289
+    },
+    {
+      "epoch": 0.26351658337119493,
+      "grad_norm": 0.2711164984704392,
+      "learning_rate": 2.6438971634393416e-05,
+      "loss": 0.5534,
+      "step": 290
+    },
+    {
+      "epoch": 0.2644252612448887,
+      "grad_norm": 0.26684683603216064,
+      "learning_rate": 2.6409749506090456e-05,
+      "loss": 0.5361,
+      "step": 291
+    },
+    {
+      "epoch": 0.26533393911858244,
+      "grad_norm": 0.2845466784771231,
+      "learning_rate": 2.6380424257777308e-05,
+      "loss": 0.5339,
+      "step": 292
+    },
+    {
+      "epoch": 0.26624261699227625,
+      "grad_norm": 0.2731604690037556,
+      "learning_rate": 2.6350996154492207e-05,
+      "loss": 0.5042,
+      "step": 293
+    },
+    {
+      "epoch": 0.26715129486597,
+      "grad_norm": 0.3036904378905144,
+      "learning_rate": 2.6321465462202985e-05,
+      "loss": 0.5345,
+      "step": 294
+    },
+    {
+      "epoch": 0.2680599727396638,
+      "grad_norm": 0.24972173802456996,
+      "learning_rate": 2.6291832447804658e-05,
+      "loss": 0.521,
+      "step": 295
+    },
+    {
+      "epoch": 0.26896865061335756,
+      "grad_norm": 0.3007524094442608,
+      "learning_rate": 2.6262097379117015e-05,
+      "loss": 0.5585,
+      "step": 296
+    },
+    {
+      "epoch": 0.2698773284870513,
+      "grad_norm": 0.2362622844963897,
+      "learning_rate": 2.623226052488221e-05,
+      "loss": 0.5234,
+      "step": 297
+    },
+    {
+      "epoch": 0.2707860063607451,
+      "grad_norm": 0.23694975837443996,
+      "learning_rate": 2.620232215476231e-05,
+      "loss": 0.5302,
+      "step": 298
+    },
+    {
+      "epoch": 0.2716946842344389,
+      "grad_norm": 0.2549336943473087,
+      "learning_rate": 2.6172282539336883e-05,
+      "loss": 0.5046,
+      "step": 299
+    },
+    {
+      "epoch": 0.2726033621081327,
+      "grad_norm": 0.27313340626086924,
+      "learning_rate": 2.6142141950100533e-05,
+      "loss": 0.5341,
+      "step": 300
+    },
+    {
+      "epoch": 0.27351203998182644,
+      "grad_norm": 0.24090945244526518,
+      "learning_rate": 2.6111900659460455e-05,
+      "loss": 0.5174,
+      "step": 301
+    },
+    {
+      "epoch": 0.2744207178555202,
+      "grad_norm": 0.2577327230433516,
+      "learning_rate": 2.6081558940733974e-05,
+      "loss": 0.5328,
+      "step": 302
+    },
+    {
+      "epoch": 0.275329395729214,
+      "grad_norm": 0.2751050941709027,
+      "learning_rate": 2.6051117068146073e-05,
+      "loss": 0.5304,
+      "step": 303
+    },
+    {
+      "epoch": 0.27623807360290775,
+      "grad_norm": 0.24804858681864422,
+      "learning_rate": 2.6020575316826912e-05,
+      "loss": 0.5271,
+      "step": 304
+    },
+    {
+      "epoch": 0.27714675147660156,
+      "grad_norm": 0.2661161507676517,
+      "learning_rate": 2.598993396280934e-05,
+      "loss": 0.5082,
+      "step": 305
+    },
+    {
+      "epoch": 0.2780554293502953,
+      "grad_norm": 0.2604861986147373,
+      "learning_rate": 2.595919328302641e-05,
+      "loss": 0.5237,
+      "step": 306
+    },
+    {
+      "epoch": 0.2789641072239891,
+      "grad_norm": 0.26623040665758096,
+      "learning_rate": 2.5928353555308873e-05,
+      "loss": 0.521,
+      "step": 307
+    },
+    {
+      "epoch": 0.2798727850976829,
+      "grad_norm": 0.2545017248380541,
+      "learning_rate": 2.5897415058382645e-05,
+      "loss": 0.5244,
+      "step": 308
+    },
+    {
+      "epoch": 0.2807814629713766,
+      "grad_norm": 0.2977913139309013,
+      "learning_rate": 2.5866378071866338e-05,
+      "loss": 0.536,
+      "step": 309
+    },
+    {
+      "epoch": 0.28169014084507044,
+      "grad_norm": 0.2640290725076467,
+      "learning_rate": 2.5835242876268667e-05,
+      "loss": 0.5169,
+      "step": 310
+    },
+    {
+      "epoch": 0.2825988187187642,
+      "grad_norm": 0.272710612040365,
+      "learning_rate": 2.5804009752985975e-05,
+      "loss": 0.5013,
+      "step": 311
+    },
+    {
+      "epoch": 0.283507496592458,
+      "grad_norm": 0.25625692372703074,
+      "learning_rate": 2.577267898429966e-05,
+      "loss": 0.5268,
+      "step": 312
+    },
+    {
+      "epoch": 0.28441617446615175,
+      "grad_norm": 0.2899485136934706,
+      "learning_rate": 2.574125085337361e-05,
+      "loss": 0.5179,
+      "step": 313
+    },
+    {
+      "epoch": 0.2853248523398455,
+      "grad_norm": 0.2719023563096517,
+      "learning_rate": 2.570972564425169e-05,
+      "loss": 0.5258,
+      "step": 314
+    },
+    {
+      "epoch": 0.2862335302135393,
+      "grad_norm": 0.27836053706475294,
+      "learning_rate": 2.5678103641855122e-05,
+      "loss": 0.5257,
+      "step": 315
+    },
+    {
+      "epoch": 0.28714220808723306,
+      "grad_norm": 0.26623698724894534,
+      "learning_rate": 2.564638513197995e-05,
+      "loss": 0.5325,
+      "step": 316
+    },
+    {
+      "epoch": 0.2880508859609269,
+      "grad_norm": 0.27117077475624796,
+      "learning_rate": 2.5614570401294434e-05,
+      "loss": 0.5123,
+      "step": 317
+    },
+    {
+      "epoch": 0.2889595638346206,
+      "grad_norm": 0.27926276040984455,
+      "learning_rate": 2.5582659737336468e-05,
+      "loss": 0.5344,
+      "step": 318
+    },
+    {
+      "epoch": 0.2898682417083144,
+      "grad_norm": 0.2693820558084933,
+      "learning_rate": 2.555065342851098e-05,
+      "loss": 0.5381,
+      "step": 319
+    },
+    {
+      "epoch": 0.2907769195820082,
+      "grad_norm": 0.24510109791420934,
+      "learning_rate": 2.5518551764087326e-05,
+      "loss": 0.5213,
+      "step": 320
+    },
+    {
+      "epoch": 0.29168559745570194,
+      "grad_norm": 0.22664216238654272,
+      "learning_rate": 2.5486355034196686e-05,
+      "loss": 0.532,
+      "step": 321
+    },
+    {
+      "epoch": 0.29259427532939575,
+      "grad_norm": 0.26225099787887585,
+      "learning_rate": 2.5454063529829405e-05,
+      "loss": 0.5343,
+      "step": 322
+    },
+    {
+      "epoch": 0.2935029532030895,
+      "grad_norm": 0.27572506715780165,
+      "learning_rate": 2.5421677542832406e-05,
+      "loss": 0.5238,
+      "step": 323
+    },
+    {
+      "epoch": 0.29441163107678325,
+      "grad_norm": 0.2400899275047973,
+      "learning_rate": 2.538919736590654e-05,
+      "loss": 0.517,
+      "step": 324
+    },
+    {
+      "epoch": 0.29532030895047706,
+      "grad_norm": 0.249003004458264,
+      "learning_rate": 2.5356623292603923e-05,
+      "loss": 0.5425,
+      "step": 325
+    },
+    {
+      "epoch": 0.2962289868241708,
+      "grad_norm": 0.27495906745966536,
+      "learning_rate": 2.53239556173253e-05,
+      "loss": 0.5268,
+      "step": 326
+    },
+    {
+      "epoch": 0.2971376646978646,
+      "grad_norm": 0.2932613052572052,
+      "learning_rate": 2.529119463531739e-05,
+      "loss": 0.5304,
+      "step": 327
+    },
+    {
+      "epoch": 0.2980463425715584,
+      "grad_norm": 0.4467908726632037,
+      "learning_rate": 2.525834064267019e-05,
+      "loss": 0.5188,
+      "step": 328
+    },
+    {
+      "epoch": 0.2989550204452522,
+      "grad_norm": 0.2563779834024989,
+      "learning_rate": 2.5225393936314328e-05,
+      "loss": 0.5217,
+      "step": 329
+    },
+    {
+      "epoch": 0.29986369831894594,
+      "grad_norm": 0.2828739499393757,
+      "learning_rate": 2.519235481401837e-05,
+      "loss": 0.5417,
+      "step": 330
+    },
+    {
+      "epoch": 0.3007723761926397,
+      "grad_norm": 0.29573580299114555,
+      "learning_rate": 2.5159223574386117e-05,
+      "loss": 0.5336,
+      "step": 331
+    },
+    {
+      "epoch": 0.3016810540663335,
+      "grad_norm": 0.2612869803912411,
+      "learning_rate": 2.512600051685392e-05,
+      "loss": 0.5066,
+      "step": 332
+    },
+    {
+      "epoch": 0.30258973194002725,
+      "grad_norm": 0.24637368673414986,
+      "learning_rate": 2.509268594168797e-05,
+      "loss": 0.5121,
+      "step": 333
+    },
+    {
+      "epoch": 0.30349840981372106,
+      "grad_norm": 0.2593406096692407,
+      "learning_rate": 2.5059280149981598e-05,
+      "loss": 0.5173,
+      "step": 334
+    },
+    {
+      "epoch": 0.3044070876874148,
+      "grad_norm": 0.28821198005365534,
+      "learning_rate": 2.5025783443652504e-05,
+      "loss": 0.5307,
+      "step": 335
+    },
+    {
+      "epoch": 0.30531576556110857,
+      "grad_norm": 0.24808902238741312,
+      "learning_rate": 2.49921961254401e-05,
+      "loss": 0.5319,
+      "step": 336
+    },
+    {
+      "epoch": 0.3062244434348024,
+      "grad_norm": 0.23581257303037423,
+      "learning_rate": 2.4958518498902725e-05,
+      "loss": 0.5056,
+      "step": 337
+    },
+    {
+      "epoch": 0.3071331213084961,
+      "grad_norm": 0.30488588123260907,
+      "learning_rate": 2.4924750868414907e-05,
+      "loss": 0.5264,
+      "step": 338
+    },
+    {
+      "epoch": 0.30804179918218993,
+      "grad_norm": 0.2332850709261921,
+      "learning_rate": 2.489089353916463e-05,
+      "loss": 0.5085,
+      "step": 339
+    },
+    {
+      "epoch": 0.3089504770558837,
+      "grad_norm": 0.2556648236725686,
+      "learning_rate": 2.4856946817150555e-05,
+      "loss": 0.5317,
+      "step": 340
+    },
+    {
+      "epoch": 0.30985915492957744,
+      "grad_norm": 0.25567477216423434,
+      "learning_rate": 2.482291100917928e-05,
+      "loss": 0.5161,
+      "step": 341
+    },
+    {
+      "epoch": 0.31076783280327125,
+      "grad_norm": 0.27072507173981075,
+      "learning_rate": 2.478878642286253e-05,
+      "loss": 0.5321,
+      "step": 342
+    },
+    {
+      "epoch": 0.311676510676965,
+      "grad_norm": 0.2605115373745453,
+      "learning_rate": 2.4754573366614418e-05,
+      "loss": 0.5256,
+      "step": 343
+    },
+    {
+      "epoch": 0.3125851885506588,
+      "grad_norm": 0.24858959605676012,
+      "learning_rate": 2.4720272149648632e-05,
+      "loss": 0.5386,
+      "step": 344
+    },
+    {
+      "epoch": 0.31349386642435256,
+      "grad_norm": 0.24688132317645303,
+      "learning_rate": 2.4685883081975646e-05,
+      "loss": 0.4936,
+      "step": 345
+    },
+    {
+      "epoch": 0.3144025442980463,
+      "grad_norm": 0.24959974360358334,
+      "learning_rate": 2.465140647439991e-05,
+      "loss": 0.5247,
+      "step": 346
+    },
+    {
+      "epoch": 0.3153112221717401,
+      "grad_norm": 0.25991911172601784,
+      "learning_rate": 2.4616842638517064e-05,
+      "loss": 0.5362,
+      "step": 347
+    },
+    {
+      "epoch": 0.3162199000454339,
+      "grad_norm": 0.23071672077330374,
+      "learning_rate": 2.4582191886711093e-05,
+      "loss": 0.5209,
+      "step": 348
+    },
+    {
+      "epoch": 0.3171285779191277,
+      "grad_norm": 0.2929778093811859,
+      "learning_rate": 2.4547454532151535e-05,
+      "loss": 0.5096,
+      "step": 349
+    },
+    {
+      "epoch": 0.31803725579282144,
+      "grad_norm": 0.2389578721549624,
+      "learning_rate": 2.4512630888790623e-05,
+      "loss": 0.5139,
+      "step": 350
+    },
+    {
+      "epoch": 0.31894593366651525,
+      "grad_norm": 0.26314034132893005,
+      "learning_rate": 2.447772127136046e-05,
+      "loss": 0.5227,
+      "step": 351
+    },
+    {
+      "epoch": 0.319854611540209,
+      "grad_norm": 0.2875398798502879,
+      "learning_rate": 2.4442725995370172e-05,
+      "loss": 0.5293,
+      "step": 352
+    },
+    {
+      "epoch": 0.32076328941390275,
+      "grad_norm": 0.2419783839135977,
+      "learning_rate": 2.4407645377103056e-05,
+      "loss": 0.5063,
+      "step": 353
+    },
+    {
+      "epoch": 0.32167196728759656,
+      "grad_norm": 0.247103751631837,
+      "learning_rate": 2.4372479733613728e-05,
+      "loss": 0.5351,
+      "step": 354
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 0.2659228516335401,
+      "learning_rate": 2.4337229382725245e-05,
+      "loss": 0.5115,
+      "step": 355
+    },
+    {
+      "epoch": 0.3234893230349841,
+      "grad_norm": 0.2831531321069596,
+      "learning_rate": 2.430189464302625e-05,
+      "loss": 0.5331,
+      "step": 356
+    },
+    {
+      "epoch": 0.3243980009086779,
+      "grad_norm": 0.24482732390316203,
+      "learning_rate": 2.4266475833868074e-05,
+      "loss": 0.5233,
+      "step": 357
+    },
+    {
+      "epoch": 0.3253066787823716,
+      "grad_norm": 0.23977791537507986,
+      "learning_rate": 2.4230973275361858e-05,
+      "loss": 0.5355,
+      "step": 358
+    },
+    {
+      "epoch": 0.32621535665606544,
+      "grad_norm": 0.24718631252268783,
+      "learning_rate": 2.4195387288375667e-05,
+      "loss": 0.526,
+      "step": 359
+    },
+    {
+      "epoch": 0.3271240345297592,
+      "grad_norm": 0.23087062790002885,
+      "learning_rate": 2.4159718194531573e-05,
+      "loss": 0.534,
+      "step": 360
+    },
+    {
+      "epoch": 0.328032712403453,
+      "grad_norm": 0.24621497452205773,
+      "learning_rate": 2.4123966316202768e-05,
+      "loss": 0.5249,
+      "step": 361
+    },
+    {
+      "epoch": 0.32894139027714675,
+      "grad_norm": 0.23192899573599338,
+      "learning_rate": 2.4088131976510633e-05,
+      "loss": 0.5029,
+      "step": 362
+    },
+    {
+      "epoch": 0.3298500681508405,
+      "grad_norm": 0.26992995982330875,
+      "learning_rate": 2.4052215499321837e-05,
+      "loss": 0.537,
+      "step": 363
+    },
+    {
+      "epoch": 0.3307587460245343,
+      "grad_norm": 0.24505808341102853,
+      "learning_rate": 2.4016217209245377e-05,
+      "loss": 0.5144,
+      "step": 364
+    },
+    {
+      "epoch": 0.33166742389822806,
+      "grad_norm": 0.2593264924319726,
+      "learning_rate": 2.398013743162969e-05,
+      "loss": 0.5136,
+      "step": 365
+    },
+    {
+      "epoch": 0.3325761017719219,
+      "grad_norm": 0.26942263570642233,
+      "learning_rate": 2.3943976492559675e-05,
+      "loss": 0.526,
+      "step": 366
+    },
+    {
+      "epoch": 0.3334847796456156,
+      "grad_norm": 0.25903401709622187,
+      "learning_rate": 2.3907734718853756e-05,
+      "loss": 0.5274,
+      "step": 367
+    },
+    {
+      "epoch": 0.3343934575193094,
+      "grad_norm": 0.2629598083863886,
+      "learning_rate": 2.3871412438060933e-05,
+      "loss": 0.5229,
+      "step": 368
+    },
+    {
+      "epoch": 0.3353021353930032,
+      "grad_norm": 0.2661133422715312,
+      "learning_rate": 2.3835009978457833e-05,
+      "loss": 0.5508,
+      "step": 369
+    },
+    {
+      "epoch": 0.33621081326669694,
+      "grad_norm": 0.266262476527216,
+      "learning_rate": 2.3798527669045714e-05,
+      "loss": 0.4896,
+      "step": 370
+    },
+    {
+      "epoch": 0.33711949114039075,
+      "grad_norm": 0.2552822951159819,
+      "learning_rate": 2.3761965839547515e-05,
+      "loss": 0.5043,
+      "step": 371
+    },
+    {
+      "epoch": 0.3380281690140845,
+      "grad_norm": 0.26247510465009727,
+      "learning_rate": 2.3725324820404864e-05,
+      "loss": 0.5212,
+      "step": 372
+    },
+    {
+      "epoch": 0.3389368468877783,
+      "grad_norm": 0.26377209436914206,
+      "learning_rate": 2.3688604942775104e-05,
+      "loss": 0.5003,
+      "step": 373
+    },
+    {
+      "epoch": 0.33984552476147206,
+      "grad_norm": 0.2866216584456117,
+      "learning_rate": 2.3651806538528277e-05,
+      "loss": 0.5394,
+      "step": 374
+    },
+    {
+      "epoch": 0.3407542026351658,
+      "grad_norm": 0.321213963036692,
+      "learning_rate": 2.3614929940244155e-05,
+      "loss": 0.5228,
+      "step": 375
+    },
+    {
+      "epoch": 0.3416628805088596,
+      "grad_norm": 0.27994833135501873,
+      "learning_rate": 2.3577975481209214e-05,
+      "loss": 0.5447,
+      "step": 376
+    },
+    {
+      "epoch": 0.3425715583825534,
+      "grad_norm": 0.26280605731029355,
+      "learning_rate": 2.3540943495413624e-05,
+      "loss": 0.5192,
+      "step": 377
+    },
+    {
+      "epoch": 0.3434802362562472,
+      "grad_norm": 0.28019047768430555,
+      "learning_rate": 2.3503834317548227e-05,
+      "loss": 0.5033,
+      "step": 378
+    },
+    {
+      "epoch": 0.34438891412994094,
+      "grad_norm": 0.2507309672329866,
+      "learning_rate": 2.3466648283001542e-05,
+      "loss": 0.5396,
+      "step": 379
+    },
+    {
+      "epoch": 0.3452975920036347,
+      "grad_norm": 0.29912678040761936,
+      "learning_rate": 2.3429385727856674e-05,
+      "loss": 0.5248,
+      "step": 380
+    },
+    {
+      "epoch": 0.3462062698773285,
+      "grad_norm": 0.2637675471489468,
+      "learning_rate": 2.3392046988888345e-05,
+      "loss": 0.5139,
+      "step": 381
+    },
+    {
+      "epoch": 0.34711494775102225,
+      "grad_norm": 0.25935799301918555,
+      "learning_rate": 2.33546324035598e-05,
+      "loss": 0.5367,
+      "step": 382
+    },
+    {
+      "epoch": 0.34802362562471606,
+      "grad_norm": 0.3066244700426474,
+      "learning_rate": 2.331714231001978e-05,
+      "loss": 0.5009,
+      "step": 383
+    },
+    {
+      "epoch": 0.3489323034984098,
+      "grad_norm": 0.2992183911705901,
+      "learning_rate": 2.3279577047099457e-05,
+      "loss": 0.512,
+      "step": 384
+    },
+    {
+      "epoch": 0.34984098137210357,
+      "grad_norm": 0.2695589550918868,
+      "learning_rate": 2.3241936954309382e-05,
+      "loss": 0.5311,
+      "step": 385
+    },
+    {
+      "epoch": 0.3507496592457974,
+      "grad_norm": 0.23292795209077574,
+      "learning_rate": 2.320422237183641e-05,
+      "loss": 0.5101,
+      "step": 386
+    },
+    {
+      "epoch": 0.3516583371194911,
+      "grad_norm": 0.239637841319824,
+      "learning_rate": 2.3166433640540602e-05,
+      "loss": 0.5011,
+      "step": 387
+    },
+    {
+      "epoch": 0.35256701499318494,
+      "grad_norm": 0.29108219617293946,
+      "learning_rate": 2.3128571101952207e-05,
+      "loss": 0.5117,
+      "step": 388
+    },
+    {
+      "epoch": 0.3534756928668787,
+      "grad_norm": 0.259703792122574,
+      "learning_rate": 2.3090635098268505e-05,
+      "loss": 0.5,
+      "step": 389
+    },
+    {
+      "epoch": 0.35438437074057244,
+      "grad_norm": 0.2379621716490283,
+      "learning_rate": 2.305262597235076e-05,
+      "loss": 0.5228,
+      "step": 390
+    },
+    {
+      "epoch": 0.35529304861426625,
+      "grad_norm": 0.26001892708811136,
+      "learning_rate": 2.3014544067721096e-05,
+      "loss": 0.5176,
+      "step": 391
+    },
+    {
+      "epoch": 0.35620172648796,
+      "grad_norm": 0.24496114487545176,
+      "learning_rate": 2.297638972855942e-05,
+      "loss": 0.5331,
+      "step": 392
+    },
+    {
+      "epoch": 0.3571104043616538,
+      "grad_norm": 0.2518685059909613,
+      "learning_rate": 2.293816329970027e-05,
+      "loss": 0.5063,
+      "step": 393
+    },
+    {
+      "epoch": 0.35801908223534756,
+      "grad_norm": 0.250752775705928,
+      "learning_rate": 2.289986512662974e-05,
+      "loss": 0.5397,
+      "step": 394
+    },
+    {
+      "epoch": 0.35892776010904137,
+      "grad_norm": 0.2701120016725626,
+      "learning_rate": 2.2861495555482337e-05,
+      "loss": 0.5233,
+      "step": 395
+    },
+    {
+      "epoch": 0.3598364379827351,
+      "grad_norm": 0.2638105659756278,
+      "learning_rate": 2.282305493303785e-05,
+      "loss": 0.5294,
+      "step": 396
+    },
+    {
+      "epoch": 0.3607451158564289,
+      "grad_norm": 0.24041324098221906,
+      "learning_rate": 2.2784543606718227e-05,
+      "loss": 0.518,
+      "step": 397
+    },
+    {
+      "epoch": 0.3616537937301227,
+      "grad_norm": 0.259602640344262,
+      "learning_rate": 2.274596192458443e-05,
+      "loss": 0.5234,
+      "step": 398
+    },
+    {
+      "epoch": 0.36256247160381644,
+      "grad_norm": 0.23222802536767737,
+      "learning_rate": 2.2707310235333284e-05,
+      "loss": 0.4881,
+      "step": 399
+    },
+    {
+      "epoch": 0.36347114947751025,
+      "grad_norm": 0.2443445481061016,
+      "learning_rate": 2.266858888829433e-05,
+      "loss": 0.5235,
+      "step": 400
+    },
+    {
+      "epoch": 0.364379827351204,
+      "grad_norm": 0.26006596580077374,
+      "learning_rate": 2.2629798233426677e-05,
+      "loss": 0.5028,
+      "step": 401
+    },
+    {
+      "epoch": 0.36528850522489775,
+      "grad_norm": 0.22354431116478168,
+      "learning_rate": 2.259093862131582e-05,
+      "loss": 0.5173,
+      "step": 402
+    },
+    {
+      "epoch": 0.36619718309859156,
+      "grad_norm": 0.22101715081783416,
+      "learning_rate": 2.2552010403170486e-05,
+      "loss": 0.5164,
+      "step": 403
+    },
+    {
+      "epoch": 0.3671058609722853,
+      "grad_norm": 0.25855149226438745,
+      "learning_rate": 2.2513013930819452e-05,
+      "loss": 0.524,
+      "step": 404
+    },
+    {
+      "epoch": 0.3680145388459791,
+      "grad_norm": 0.2500646351802012,
+      "learning_rate": 2.247394955670838e-05,
+      "loss": 0.5099,
+      "step": 405
+    },
+    {
+      "epoch": 0.3689232167196729,
+      "grad_norm": 0.22785523897152013,
+      "learning_rate": 2.243481763389661e-05,
+      "loss": 0.5351,
+      "step": 406
+    },
+    {
+      "epoch": 0.36983189459336663,
+      "grad_norm": 0.2742012018148663,
+      "learning_rate": 2.239561851605397e-05,
+      "loss": 0.5294,
+      "step": 407
+    },
+    {
+      "epoch": 0.37074057246706044,
+      "grad_norm": 0.2917409445353074,
+      "learning_rate": 2.2356352557457624e-05,
+      "loss": 0.5245,
+      "step": 408
+    },
+    {
+      "epoch": 0.3716492503407542,
+      "grad_norm": 0.24786694428623812,
+      "learning_rate": 2.2317020112988792e-05,
+      "loss": 0.5141,
+      "step": 409
+    },
+    {
+      "epoch": 0.372557928214448,
+      "grad_norm": 0.23835087667090923,
+      "learning_rate": 2.2277621538129612e-05,
+      "loss": 0.4942,
+      "step": 410
+    },
+    {
+      "epoch": 0.37346660608814175,
+      "grad_norm": 0.24847073099295497,
+      "learning_rate": 2.2238157188959893e-05,
+      "loss": 0.515,
+      "step": 411
+    },
+    {
+      "epoch": 0.3743752839618355,
+      "grad_norm": 0.21988501584629094,
+      "learning_rate": 2.2198627422153908e-05,
+      "loss": 0.5312,
+      "step": 412
+    },
+    {
+      "epoch": 0.3752839618355293,
+      "grad_norm": 0.24762672440554007,
+      "learning_rate": 2.2159032594977165e-05,
+      "loss": 0.5197,
+      "step": 413
+    },
+    {
+      "epoch": 0.37619263970922306,
+      "grad_norm": 0.23443065509690622,
+      "learning_rate": 2.211937306528318e-05,
+      "loss": 0.5075,
+      "step": 414
+    },
+    {
+      "epoch": 0.3771013175829169,
+      "grad_norm": 0.27862711744360735,
+      "learning_rate": 2.2079649191510248e-05,
+      "loss": 0.4837,
+      "step": 415
+    },
+    {
+      "epoch": 0.3780099954566106,
+      "grad_norm": 0.23994539916874946,
+      "learning_rate": 2.203986133267818e-05,
+      "loss": 0.5129,
+      "step": 416
+    },
+    {
+      "epoch": 0.37891867333030443,
+      "grad_norm": 0.2321097118020647,
+      "learning_rate": 2.2000009848385107e-05,
+      "loss": 0.5113,
+      "step": 417
+    },
+    {
+      "epoch": 0.3798273512039982,
+      "grad_norm": 0.24702140514455406,
+      "learning_rate": 2.196009509880417e-05,
+      "loss": 0.5276,
+      "step": 418
+    },
+    {
+      "epoch": 0.38073602907769194,
+      "grad_norm": 0.31011337504149433,
+      "learning_rate": 2.1920117444680317e-05,
+      "loss": 0.5235,
+      "step": 419
+    },
+    {
+      "epoch": 0.38164470695138575,
+      "grad_norm": 0.25024867689118563,
+      "learning_rate": 2.1880077247327008e-05,
+      "loss": 0.5319,
+      "step": 420
+    },
+    {
+      "epoch": 0.3825533848250795,
+      "grad_norm": 0.2657215152096355,
+      "learning_rate": 2.1839974868622956e-05,
+      "loss": 0.5178,
+      "step": 421
+    },
+    {
+      "epoch": 0.3834620626987733,
+      "grad_norm": 0.2581802279830013,
+      "learning_rate": 2.1799810671008887e-05,
+      "loss": 0.5078,
+      "step": 422
+    },
+    {
+      "epoch": 0.38437074057246706,
+      "grad_norm": 0.2528887013846824,
+      "learning_rate": 2.17595850174842e-05,
+      "loss": 0.5234,
+      "step": 423
+    },
+    {
+      "epoch": 0.3852794184461608,
+      "grad_norm": 0.238989778351192,
+      "learning_rate": 2.1719298271603766e-05,
+      "loss": 0.5121,
+      "step": 424
+    },
+    {
+      "epoch": 0.3861880963198546,
+      "grad_norm": 0.2643788344384685,
+      "learning_rate": 2.1678950797474578e-05,
+      "loss": 0.4956,
+      "step": 425
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 0.24554204369119242,
+      "learning_rate": 2.1638542959752485e-05,
+      "loss": 0.5413,
+      "step": 426
+    },
+    {
+      "epoch": 0.3880054520672422,
+      "grad_norm": 0.23118345256737963,
+      "learning_rate": 2.1598075123638903e-05,
+      "loss": 0.5137,
+      "step": 427
+    },
+    {
+      "epoch": 0.38891412994093594,
+      "grad_norm": 0.25400418452076406,
+      "learning_rate": 2.1557547654877506e-05,
+      "loss": 0.508,
+      "step": 428
+    },
+    {
+      "epoch": 0.3898228078146297,
+      "grad_norm": 0.22797693534084318,
+      "learning_rate": 2.1516960919750914e-05,
+      "loss": 0.5022,
+      "step": 429
+    },
+    {
+      "epoch": 0.3907314856883235,
+      "grad_norm": 0.23059282438934506,
+      "learning_rate": 2.1476315285077393e-05,
+      "loss": 0.4997,
+      "step": 430
+    },
+    {
+      "epoch": 0.39164016356201725,
+      "grad_norm": 0.24684953496844692,
+      "learning_rate": 2.1435611118207546e-05,
+      "loss": 0.524,
+      "step": 431
+    },
+    {
+      "epoch": 0.39254884143571106,
+      "grad_norm": 0.24358371474217685,
+      "learning_rate": 2.1394848787020963e-05,
+      "loss": 0.4959,
+      "step": 432
+    },
+    {
+      "epoch": 0.3934575193094048,
+      "grad_norm": 0.24009597119673534,
+      "learning_rate": 2.1354028659922938e-05,
+      "loss": 0.5065,
+      "step": 433
+    },
+    {
+      "epoch": 0.39436619718309857,
+      "grad_norm": 0.21258467054067773,
+      "learning_rate": 2.13131511058411e-05,
+      "loss": 0.5027,
+      "step": 434
+    },
+    {
+      "epoch": 0.3952748750567924,
+      "grad_norm": 0.2332547982597397,
+      "learning_rate": 2.1272216494222108e-05,
+      "loss": 0.521,
+      "step": 435
+    },
+    {
+      "epoch": 0.3961835529304861,
+      "grad_norm": 0.26527023398234467,
+      "learning_rate": 2.12312251950283e-05,
+      "loss": 0.51,
+      "step": 436
+    },
+    {
+      "epoch": 0.39709223080417994,
+      "grad_norm": 0.24686604477814575,
+      "learning_rate": 2.1190177578734336e-05,
+      "loss": 0.5217,
+      "step": 437
+    },
+    {
+      "epoch": 0.3980009086778737,
+      "grad_norm": 0.23200522771280188,
+      "learning_rate": 2.114907401632389e-05,
+      "loss": 0.5182,
+      "step": 438
+    },
+    {
+      "epoch": 0.3989095865515675,
+      "grad_norm": 0.22649180511207603,
+      "learning_rate": 2.1107914879286237e-05,
+      "loss": 0.5103,
+      "step": 439
+    },
+    {
+      "epoch": 0.39981826442526125,
+      "grad_norm": 0.2885194149892351,
+      "learning_rate": 2.1066700539612964e-05,
+      "loss": 0.5293,
+      "step": 440
+    },
+    {
+      "epoch": 0.400726942298955,
+      "grad_norm": 0.22579048097840532,
+      "learning_rate": 2.1025431369794546e-05,
+      "loss": 0.5257,
+      "step": 441
+    },
+    {
+      "epoch": 0.4016356201726488,
+      "grad_norm": 0.27513365175752,
+      "learning_rate": 2.0984107742817016e-05,
+      "loss": 0.5401,
+      "step": 442
+    },
+    {
+      "epoch": 0.40254429804634256,
+      "grad_norm": 0.2884279516810869,
+      "learning_rate": 2.0942730032158586e-05,
+      "loss": 0.5227,
+      "step": 443
+    },
+    {
+      "epoch": 0.40345297592003637,
+      "grad_norm": 0.276114207969109,
+      "learning_rate": 2.0901298611786274e-05,
+      "loss": 0.5297,
+      "step": 444
+    },
+    {
+      "epoch": 0.4043616537937301,
+      "grad_norm": 0.254987448745979,
+      "learning_rate": 2.0859813856152513e-05,
+      "loss": 0.508,
+      "step": 445
+    },
+    {
+      "epoch": 0.4052703316674239,
+      "grad_norm": 0.2700548502662214,
+      "learning_rate": 2.081827614019177e-05,
+      "loss": 0.5056,
+      "step": 446
+    },
+    {
+      "epoch": 0.4061790095411177,
+      "grad_norm": 0.26239233598339107,
+      "learning_rate": 2.077668583931718e-05,
+      "loss": 0.5185,
+      "step": 447
+    },
+    {
+      "epoch": 0.40708768741481144,
+      "grad_norm": 0.21565404347135067,
+      "learning_rate": 2.0735043329417115e-05,
+      "loss": 0.4941,
+      "step": 448
+    },
+    {
+      "epoch": 0.40799636528850525,
+      "grad_norm": 0.2728801118505973,
+      "learning_rate": 2.0693348986851817e-05,
+      "loss": 0.512,
+      "step": 449
+    },
+    {
+      "epoch": 0.408905043162199,
+      "grad_norm": 0.2754327035303729,
+      "learning_rate": 2.0651603188449976e-05,
+      "loss": 0.5139,
+      "step": 450
+    },
+    {
+      "epoch": 0.40981372103589275,
+      "grad_norm": 0.2655235441029784,
+      "learning_rate": 2.0609806311505345e-05,
+      "loss": 0.4772,
+      "step": 451
+    },
+    {
+      "epoch": 0.41072239890958656,
+      "grad_norm": 0.2440498900832255,
+      "learning_rate": 2.056795873377331e-05,
+      "loss": 0.4951,
+      "step": 452
+    },
+    {
+      "epoch": 0.4116310767832803,
+      "grad_norm": 0.2571323059977,
+      "learning_rate": 2.0526060833467495e-05,
+      "loss": 0.525,
+      "step": 453
+    },
+    {
+      "epoch": 0.4125397546569741,
+      "grad_norm": 0.24645495086607,
+      "learning_rate": 2.0484112989256327e-05,
+      "loss": 0.5259,
+      "step": 454
+    },
+    {
+      "epoch": 0.4134484325306679,
+      "grad_norm": 0.262434196340489,
+      "learning_rate": 2.0442115580259615e-05,
+      "loss": 0.5268,
+      "step": 455
+    },
+    {
+      "epoch": 0.41435711040436163,
+      "grad_norm": 0.24807419069104444,
+      "learning_rate": 2.0400068986045142e-05,
+      "loss": 0.5177,
+      "step": 456
+    },
+    {
+      "epoch": 0.41526578827805544,
+      "grad_norm": 0.23338885931168596,
+      "learning_rate": 2.035797358662521e-05,
+      "loss": 0.5211,
+      "step": 457
+    },
+    {
+      "epoch": 0.4161744661517492,
+      "grad_norm": 0.23650378253272059,
+      "learning_rate": 2.0315829762453223e-05,
+      "loss": 0.4978,
+      "step": 458
+    },
+    {
+      "epoch": 0.417083144025443,
+      "grad_norm": 0.2372961976074164,
+      "learning_rate": 2.0273637894420236e-05,
+      "loss": 0.5214,
+      "step": 459
+    },
+    {
+      "epoch": 0.41799182189913675,
+      "grad_norm": 0.2677513909886547,
+      "learning_rate": 2.023139836385153e-05,
+      "loss": 0.5357,
+      "step": 460
+    },
+    {
+      "epoch": 0.41890049977283056,
+      "grad_norm": 0.23710027013958848,
+      "learning_rate": 2.0189111552503142e-05,
+      "loss": 0.5196,
+      "step": 461
+    },
+    {
+      "epoch": 0.4198091776465243,
+      "grad_norm": 0.21399195686356814,
+      "learning_rate": 2.014677784255844e-05,
+      "loss": 0.5083,
+      "step": 462
+    },
+    {
+      "epoch": 0.42071785552021806,
+      "grad_norm": 0.24893610234658364,
+      "learning_rate": 2.0104397616624646e-05,
+      "loss": 0.5075,
+      "step": 463
+    },
+    {
+      "epoch": 0.4216265333939119,
+      "grad_norm": 0.23670544045308084,
+      "learning_rate": 2.006197125772939e-05,
+      "loss": 0.5059,
+      "step": 464
+    },
+    {
+      "epoch": 0.4225352112676056,
+      "grad_norm": 0.26452779453858905,
+      "learning_rate": 2.0019499149317252e-05,
+      "loss": 0.5156,
+      "step": 465
+    },
+    {
+      "epoch": 0.42344388914129943,
+      "grad_norm": 0.2766916466032981,
+      "learning_rate": 1.997698167524628e-05,
+      "loss": 0.5102,
+      "step": 466
+    },
+    {
+      "epoch": 0.4243525670149932,
+      "grad_norm": 0.23222477899271163,
+      "learning_rate": 1.9934419219784544e-05,
+      "loss": 0.5002,
+      "step": 467
+    },
+    {
+      "epoch": 0.42526124488868694,
+      "grad_norm": 0.311260792933651,
+      "learning_rate": 1.9891812167606645e-05,
+      "loss": 0.5282,
+      "step": 468
+    },
+    {
+      "epoch": 0.42616992276238075,
+      "grad_norm": 0.3199547422816849,
+      "learning_rate": 1.984916090379024e-05,
+      "loss": 0.5002,
+      "step": 469
+    },
+    {
+      "epoch": 0.4270786006360745,
+      "grad_norm": 0.22312839762188058,
+      "learning_rate": 1.980646581381258e-05,
+      "loss": 0.5219,
+      "step": 470
+    },
+    {
+      "epoch": 0.4279872785097683,
+      "grad_norm": 0.2599655899356632,
+      "learning_rate": 1.976372728354699e-05,
+      "loss": 0.5177,
+      "step": 471
+    },
+    {
+      "epoch": 0.42889595638346206,
+      "grad_norm": 0.23107652252431182,
+      "learning_rate": 1.972094569925942e-05,
+      "loss": 0.5435,
+      "step": 472
+    },
+    {
+      "epoch": 0.4298046342571558,
+      "grad_norm": 0.24171392550647774,
+      "learning_rate": 1.967812144760493e-05,
+      "loss": 0.5207,
+      "step": 473
+    },
+    {
+      "epoch": 0.4307133121308496,
+      "grad_norm": 0.2345576421072366,
+      "learning_rate": 1.963525491562421e-05,
+      "loss": 0.5438,
+      "step": 474
+    },
+    {
+      "epoch": 0.4316219900045434,
+      "grad_norm": 0.2658024201064496,
+      "learning_rate": 1.959234649074007e-05,
+      "loss": 0.5297,
+      "step": 475
+    },
+    {
+      "epoch": 0.4325306678782372,
+      "grad_norm": 0.250262769604315,
+      "learning_rate": 1.954939656075394e-05,
+      "loss": 0.517,
+      "step": 476
+    },
+    {
+      "epoch": 0.43343934575193094,
+      "grad_norm": 0.25500108379604414,
+      "learning_rate": 1.950640551384237e-05,
+      "loss": 0.5094,
+      "step": 477
+    },
+    {
+      "epoch": 0.4343480236256247,
+      "grad_norm": 0.28529434407052306,
+      "learning_rate": 1.9463373738553523e-05,
+      "loss": 0.5276,
+      "step": 478
+    },
+    {
+      "epoch": 0.4352567014993185,
+      "grad_norm": 0.23120960399287802,
+      "learning_rate": 1.9420301623803666e-05,
+      "loss": 0.538,
+      "step": 479
+    },
+    {
+      "epoch": 0.43616537937301225,
+      "grad_norm": 0.25050816340583165,
+      "learning_rate": 1.9377189558873634e-05,
+      "loss": 0.5495,
+      "step": 480
+    },
+    {
+      "epoch": 0.43707405724670606,
+      "grad_norm": 0.22534575878035212,
+      "learning_rate": 1.9334037933405337e-05,
+      "loss": 0.5186,
+      "step": 481
+    },
+    {
+      "epoch": 0.4379827351203998,
+      "grad_norm": 0.22295075494230848,
+      "learning_rate": 1.9290847137398228e-05,
+      "loss": 0.5165,
+      "step": 482
+    },
+    {
+      "epoch": 0.4388914129940936,
+      "grad_norm": 0.23579161097603182,
+      "learning_rate": 1.924761756120578e-05,
+      "loss": 0.4976,
+      "step": 483
+    },
+    {
+      "epoch": 0.4398000908677874,
+      "grad_norm": 0.21832007686471547,
+      "learning_rate": 1.9204349595531954e-05,
+      "loss": 0.5218,
+      "step": 484
+    },
+    {
+      "epoch": 0.4407087687414811,
+      "grad_norm": 0.25708298869793,
+      "learning_rate": 1.916104363142767e-05,
+      "loss": 0.5062,
+      "step": 485
+    },
+    {
+      "epoch": 0.44161744661517494,
+      "grad_norm": 0.2268902295276673,
+      "learning_rate": 1.911770006028728e-05,
+      "loss": 0.5143,
+      "step": 486
+    },
+    {
+      "epoch": 0.4425261244888687,
+      "grad_norm": 0.24778267034905432,
+      "learning_rate": 1.907431927384501e-05,
+      "loss": 0.5069,
+      "step": 487
+    },
+    {
+      "epoch": 0.4434348023625625,
+      "grad_norm": 0.23357617111003176,
+      "learning_rate": 1.9030901664171457e-05,
+      "loss": 0.5085,
+      "step": 488
+    },
+    {
+      "epoch": 0.44434348023625625,
+      "grad_norm": 0.239662748648008,
+      "learning_rate": 1.8987447623669992e-05,
+      "loss": 0.5222,
+      "step": 489
+    },
+    {
+      "epoch": 0.44525215810995,
+      "grad_norm": 0.25708866983380263,
+      "learning_rate": 1.8943957545073273e-05,
+      "loss": 0.5308,
+      "step": 490
+    },
+    {
+      "epoch": 0.4461608359836438,
+      "grad_norm": 0.21783878351194902,
+      "learning_rate": 1.8900431821439644e-05,
+      "loss": 0.5251,
+      "step": 491
+    },
+    {
+      "epoch": 0.44706951385733756,
+      "grad_norm": 0.29948916904218814,
+      "learning_rate": 1.8856870846149618e-05,
+      "loss": 0.4985,
+      "step": 492
+    },
+    {
+      "epoch": 0.44797819173103137,
+      "grad_norm": 0.23300153547188593,
+      "learning_rate": 1.8813275012902307e-05,
+      "loss": 0.5163,
+      "step": 493
+    },
+    {
+      "epoch": 0.4488868696047251,
+      "grad_norm": 0.24619621895860472,
+      "learning_rate": 1.8769644715711854e-05,
+      "loss": 0.5106,
+      "step": 494
+    },
+    {
+      "epoch": 0.4497955474784189,
+      "grad_norm": 0.22818266306630752,
+      "learning_rate": 1.8725980348903898e-05,
+      "loss": 0.514,
+      "step": 495
+    },
+    {
+      "epoch": 0.4507042253521127,
+      "grad_norm": 0.22392621802844048,
+      "learning_rate": 1.8682282307111988e-05,
+      "loss": 0.5147,
+      "step": 496
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 0.22400650722344845,
+      "learning_rate": 1.8638550985274025e-05,
+      "loss": 0.5283,
+      "step": 497
+    },
+    {
+      "epoch": 0.45252158109950025,
+      "grad_norm": 0.24057049058886684,
+      "learning_rate": 1.8594786778628686e-05,
+      "loss": 0.5105,
+      "step": 498
+    },
+    {
+      "epoch": 0.453430258973194,
+      "grad_norm": 0.23834314235207676,
+      "learning_rate": 1.8550990082711868e-05,
+      "loss": 0.5142,
+      "step": 499
+    },
+    {
+      "epoch": 0.45433893684688775,
+      "grad_norm": 0.22159251074212913,
+      "learning_rate": 1.850716129335309e-05,
+      "loss": 0.5037,
+      "step": 500
+    },
+    {
+      "epoch": 0.45524761472058156,
+      "grad_norm": 0.21883306577809505,
+      "learning_rate": 1.8463300806671936e-05,
+      "loss": 0.5243,
+      "step": 501
+    },
+    {
+      "epoch": 0.4561562925942753,
+      "grad_norm": 0.25080121721482435,
+      "learning_rate": 1.8419409019074474e-05,
+      "loss": 0.5167,
+      "step": 502
+    },
+    {
+      "epoch": 0.4570649704679691,
+      "grad_norm": 0.23602150357150198,
+      "learning_rate": 1.837548632724964e-05,
+      "loss": 0.5331,
+      "step": 503
+    },
+    {
+      "epoch": 0.4579736483416629,
+      "grad_norm": 0.22751790967544427,
+      "learning_rate": 1.8331533128165713e-05,
+      "loss": 0.5092,
+      "step": 504
+    },
+    {
+      "epoch": 0.45888232621535663,
+      "grad_norm": 0.21166065633290324,
+      "learning_rate": 1.8287549819066665e-05,
+      "loss": 0.5291,
+      "step": 505
+    },
+    {
+      "epoch": 0.45979100408905044,
+      "grad_norm": 0.21303196091882118,
+      "learning_rate": 1.824353679746861e-05,
+      "loss": 0.5109,
+      "step": 506
+    },
+    {
+      "epoch": 0.4606996819627442,
+      "grad_norm": 0.2520169158737603,
+      "learning_rate": 1.8199494461156203e-05,
+      "loss": 0.5209,
+      "step": 507
+    },
+    {
+      "epoch": 0.461608359836438,
+      "grad_norm": 0.23204211684465506,
+      "learning_rate": 1.815542320817904e-05,
+      "loss": 0.5114,
+      "step": 508
+    },
+    {
+      "epoch": 0.46251703771013175,
+      "grad_norm": 0.22206027132403938,
+      "learning_rate": 1.8111323436848057e-05,
+      "loss": 0.5101,
+      "step": 509
+    },
+    {
+      "epoch": 0.46342571558382556,
+      "grad_norm": 0.2197827409132378,
+      "learning_rate": 1.8067195545731942e-05,
+      "loss": 0.5085,
+      "step": 510
+    },
+    {
+      "epoch": 0.4643343934575193,
+      "grad_norm": 0.22001178565277088,
+      "learning_rate": 1.802303993365353e-05,
+      "loss": 0.4998,
+      "step": 511
+    },
+    {
+      "epoch": 0.46524307133121307,
+      "grad_norm": 0.24609354561612468,
+      "learning_rate": 1.7978856999686182e-05,
+      "loss": 0.5178,
+      "step": 512
+    },
+    {
+      "epoch": 0.4661517492049069,
+      "grad_norm": 0.22050322371740638,
+      "learning_rate": 1.7934647143150202e-05,
+      "loss": 0.5227,
+      "step": 513
+    },
+    {
+      "epoch": 0.4670604270786006,
+      "grad_norm": 0.21073901613729007,
+      "learning_rate": 1.7890410763609212e-05,
+      "loss": 0.5006,
+      "step": 514
+    },
+    {
+      "epoch": 0.46796910495229443,
+      "grad_norm": 0.24141869359808918,
+      "learning_rate": 1.784614826086655e-05,
+      "loss": 0.5014,
+      "step": 515
+    },
+    {
+      "epoch": 0.4688777828259882,
+      "grad_norm": 0.2279765094681949,
+      "learning_rate": 1.780186003496164e-05,
+      "loss": 0.5133,
+      "step": 516
+    },
+    {
+      "epoch": 0.46978646069968194,
+      "grad_norm": 0.23855167092742002,
+      "learning_rate": 1.775754648616642e-05,
+      "loss": 0.5135,
+      "step": 517
+    },
+    {
+      "epoch": 0.47069513857337575,
+      "grad_norm": 0.21569471812715105,
+      "learning_rate": 1.771320801498165e-05,
+      "loss": 0.5272,
+      "step": 518
+    },
+    {
+      "epoch": 0.4716038164470695,
+      "grad_norm": 0.23601732409769477,
+      "learning_rate": 1.7668845022133366e-05,
+      "loss": 0.5359,
+      "step": 519
+    },
+    {
+      "epoch": 0.4725124943207633,
+      "grad_norm": 0.2421070076653461,
+      "learning_rate": 1.7624457908569226e-05,
+      "loss": 0.5151,
+      "step": 520
+    },
+    {
+      "epoch": 0.47342117219445706,
+      "grad_norm": 0.21994200464509472,
+      "learning_rate": 1.7580047075454877e-05,
+      "loss": 0.5257,
+      "step": 521
+    },
+    {
+      "epoch": 0.4743298500681508,
+      "grad_norm": 0.21413213779222634,
+      "learning_rate": 1.753561292417035e-05,
+      "loss": 0.5139,
+      "step": 522
+    },
+    {
+      "epoch": 0.4752385279418446,
+      "grad_norm": 0.2273069993329325,
+      "learning_rate": 1.7491155856306406e-05,
+      "loss": 0.5195,
+      "step": 523
+    },
+    {
+      "epoch": 0.4761472058155384,
+      "grad_norm": 0.22197392010344422,
+      "learning_rate": 1.7446676273660954e-05,
+      "loss": 0.4834,
+      "step": 524
+    },
+    {
+      "epoch": 0.4770558836892322,
+      "grad_norm": 0.21867064721147128,
+      "learning_rate": 1.7402174578235356e-05,
+      "loss": 0.5227,
+      "step": 525
+    },
+    {
+      "epoch": 0.47796456156292594,
+      "grad_norm": 0.2237126581240893,
+      "learning_rate": 1.7357651172230852e-05,
+      "loss": 0.5375,
+      "step": 526
+    },
+    {
+      "epoch": 0.4788732394366197,
+      "grad_norm": 0.23009669061440477,
+      "learning_rate": 1.731310645804489e-05,
+      "loss": 0.5411,
+      "step": 527
+    },
+    {
+      "epoch": 0.4797819173103135,
+      "grad_norm": 0.21945124564589844,
+      "learning_rate": 1.7268540838267493e-05,
+      "loss": 0.4903,
+      "step": 528
+    },
+    {
+      "epoch": 0.48069059518400725,
+      "grad_norm": 0.21735802469423854,
+      "learning_rate": 1.722395471567763e-05,
+      "loss": 0.526,
+      "step": 529
+    },
+    {
+      "epoch": 0.48159927305770106,
+      "grad_norm": 0.2259120283462735,
+      "learning_rate": 1.7179348493239582e-05,
+      "loss": 0.5099,
+      "step": 530
+    },
+    {
+      "epoch": 0.4825079509313948,
+      "grad_norm": 0.2236406553388163,
+      "learning_rate": 1.713472257409928e-05,
+      "loss": 0.5035,
+      "step": 531
+    },
+    {
+      "epoch": 0.4834166288050886,
+      "grad_norm": 0.2203648050330253,
+      "learning_rate": 1.7090077361580667e-05,
+      "loss": 0.5242,
+      "step": 532
+    },
+    {
+      "epoch": 0.4843253066787824,
+      "grad_norm": 0.21577849058700352,
+      "learning_rate": 1.7045413259182077e-05,
+      "loss": 0.5093,
+      "step": 533
+    },
+    {
+      "epoch": 0.4852339845524761,
+      "grad_norm": 0.26474820424051165,
+      "learning_rate": 1.7000730670572542e-05,
+      "loss": 0.5367,
+      "step": 534
+    },
+    {
+      "epoch": 0.48614266242616994,
+      "grad_norm": 0.21495788299510912,
+      "learning_rate": 1.695602999958819e-05,
+      "loss": 0.5144,
+      "step": 535
+    },
+    {
+      "epoch": 0.4870513402998637,
+      "grad_norm": 0.22089173778431492,
+      "learning_rate": 1.6911311650228574e-05,
+      "loss": 0.4924,
+      "step": 536
+    },
+    {
+      "epoch": 0.4879600181735575,
+      "grad_norm": 0.2276117083165815,
+      "learning_rate": 1.6866576026653006e-05,
+      "loss": 0.5011,
+      "step": 537
+    },
+    {
+      "epoch": 0.48886869604725125,
+      "grad_norm": 0.23006884025462068,
+      "learning_rate": 1.682182353317695e-05,
+      "loss": 0.5149,
+      "step": 538
+    },
+    {
+      "epoch": 0.489777373920945,
+      "grad_norm": 0.23790427840650574,
+      "learning_rate": 1.67770545742683e-05,
+      "loss": 0.5215,
+      "step": 539
+    },
+    {
+      "epoch": 0.4906860517946388,
+      "grad_norm": 0.20892374204766953,
+      "learning_rate": 1.6732269554543794e-05,
+      "loss": 0.535,
+      "step": 540
+    },
+    {
+      "epoch": 0.49159472966833256,
+      "grad_norm": 0.2160818063170021,
+      "learning_rate": 1.668746887876531e-05,
+      "loss": 0.5083,
+      "step": 541
+    },
+    {
+      "epoch": 0.4925034075420264,
+      "grad_norm": 0.20118901727019187,
+      "learning_rate": 1.6642652951836233e-05,
+      "loss": 0.4943,
+      "step": 542
+    },
+    {
+      "epoch": 0.4934120854157201,
+      "grad_norm": 0.22282897518570308,
+      "learning_rate": 1.659782217879778e-05,
+      "loss": 0.5317,
+      "step": 543
+    },
+    {
+      "epoch": 0.4943207632894139,
+      "grad_norm": 0.22131531803306617,
+      "learning_rate": 1.655297696482534e-05,
+      "loss": 0.5116,
+      "step": 544
+    },
+    {
+      "epoch": 0.4952294411631077,
+      "grad_norm": 0.2095981771825046,
+      "learning_rate": 1.6508117715224834e-05,
+      "loss": 0.5149,
+      "step": 545
+    },
+    {
+      "epoch": 0.49613811903680144,
+      "grad_norm": 0.21419702778901387,
+      "learning_rate": 1.646324483542902e-05,
+      "loss": 0.517,
+      "step": 546
+    },
+    {
+      "epoch": 0.49704679691049525,
+      "grad_norm": 0.22616880756921168,
+      "learning_rate": 1.6418358730993862e-05,
+      "loss": 0.5114,
+      "step": 547
+    },
+    {
+      "epoch": 0.497955474784189,
+      "grad_norm": 0.2184048869089771,
+      "learning_rate": 1.637345980759481e-05,
+      "loss": 0.5362,
+      "step": 548
+    },
+    {
+      "epoch": 0.49886415265788275,
+      "grad_norm": 0.22632003781828508,
+      "learning_rate": 1.6328548471023224e-05,
+      "loss": 0.4988,
+      "step": 549
+    },
+    {
+      "epoch": 0.49977283053157656,
+      "grad_norm": 0.209816073482319,
+      "learning_rate": 1.6283625127182596e-05,
+      "loss": 0.4919,
+      "step": 550
+    },
+    {
+      "epoch": 0.5006815084052704,
+      "grad_norm": 0.25442817782757327,
+      "learning_rate": 1.623869018208499e-05,
+      "loss": 0.5227,
+      "step": 551
+    },
+    {
+      "epoch": 0.5015901862789641,
+      "grad_norm": 0.24267619179624578,
+      "learning_rate": 1.6193744041847277e-05,
+      "loss": 0.5388,
+      "step": 552
+    },
+    {
+      "epoch": 0.5024988641526579,
+      "grad_norm": 0.2228324843248023,
+      "learning_rate": 1.6148787112687544e-05,
+      "loss": 0.5087,
+      "step": 553
+    },
+    {
+      "epoch": 0.5034075420263516,
+      "grad_norm": 0.3352682201561984,
+      "learning_rate": 1.6103819800921363e-05,
+      "loss": 0.5051,
+      "step": 554
+    },
+    {
+      "epoch": 0.5043162199000454,
+      "grad_norm": 0.24806511392962588,
+      "learning_rate": 1.605884251295815e-05,
+      "loss": 0.4982,
+      "step": 555
+    },
+    {
+      "epoch": 0.5052248977737392,
+      "grad_norm": 0.24952500063687508,
+      "learning_rate": 1.6013855655297498e-05,
+      "loss": 0.5014,
+      "step": 556
+    },
+    {
+      "epoch": 0.506133575647433,
+      "grad_norm": 0.21538151492096425,
+      "learning_rate": 1.5968859634525462e-05,
+      "loss": 0.5072,
+      "step": 557
+    },
+    {
+      "epoch": 0.5070422535211268,
+      "grad_norm": 0.23504846953730568,
+      "learning_rate": 1.592385485731094e-05,
+      "loss": 0.5164,
+      "step": 558
+    },
+    {
+      "epoch": 0.5079509313948205,
+      "grad_norm": 0.24076075119687024,
+      "learning_rate": 1.5878841730401945e-05,
+      "loss": 0.4991,
+      "step": 559
+    },
+    {
+      "epoch": 0.5088596092685143,
+      "grad_norm": 0.24918458142521988,
+      "learning_rate": 1.583382066062198e-05,
+      "loss": 0.5259,
+      "step": 560
+    },
+    {
+      "epoch": 0.5097682871422081,
+      "grad_norm": 0.22309906969227764,
+      "learning_rate": 1.5788792054866314e-05,
+      "loss": 0.5177,
+      "step": 561
+    },
+    {
+      "epoch": 0.5106769650159019,
+      "grad_norm": 0.22622168642080814,
+      "learning_rate": 1.5743756320098334e-05,
+      "loss": 0.5167,
+      "step": 562
+    },
+    {
+      "epoch": 0.5115856428895956,
+      "grad_norm": 0.2328779476599096,
+      "learning_rate": 1.5698713863345858e-05,
+      "loss": 0.5028,
+      "step": 563
+    },
+    {
+      "epoch": 0.5124943207632894,
+      "grad_norm": 0.22131686640647605,
+      "learning_rate": 1.565366509169745e-05,
+      "loss": 0.5243,
+      "step": 564
+    },
+    {
+      "epoch": 0.5134029986369832,
+      "grad_norm": 0.22087250704443534,
+      "learning_rate": 1.5608610412298758e-05,
+      "loss": 0.5322,
+      "step": 565
+    },
+    {
+      "epoch": 0.514311676510677,
+      "grad_norm": 0.23457527179046703,
+      "learning_rate": 1.5563550232348813e-05,
+      "loss": 0.4915,
+      "step": 566
+    },
+    {
+      "epoch": 0.5152203543843707,
+      "grad_norm": 0.22849851842588884,
+      "learning_rate": 1.5518484959096365e-05,
+      "loss": 0.5049,
+      "step": 567
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 0.22145239435825756,
+      "learning_rate": 1.54734149998362e-05,
+      "loss": 0.5063,
+      "step": 568
+    },
+    {
+      "epoch": 0.5170377101317583,
+      "grad_norm": 0.2400202215469711,
+      "learning_rate": 1.5428340761905444e-05,
+      "loss": 0.5114,
+      "step": 569
+    },
+    {
+      "epoch": 0.5179463880054521,
+      "grad_norm": 0.2493745198857758,
+      "learning_rate": 1.5383262652679914e-05,
+      "loss": 0.5155,
+      "step": 570
+    },
+    {
+      "epoch": 0.5188550658791459,
+      "grad_norm": 0.23320737861124252,
+      "learning_rate": 1.533818107957038e-05,
+      "loss": 0.5267,
+      "step": 571
+    },
+    {
+      "epoch": 0.5197637437528396,
+      "grad_norm": 0.20584929895864323,
+      "learning_rate": 1.5293096450018966e-05,
+      "loss": 0.5057,
+      "step": 572
+    },
+    {
+      "epoch": 0.5206724216265334,
+      "grad_norm": 0.23493566008214234,
+      "learning_rate": 1.524800917149538e-05,
+      "loss": 0.4921,
+      "step": 573
+    },
+    {
+      "epoch": 0.5215810995002271,
+      "grad_norm": 0.21564463992482097,
+      "learning_rate": 1.5202919651493284e-05,
+      "loss": 0.5006,
+      "step": 574
+    },
+    {
+      "epoch": 0.522489777373921,
+      "grad_norm": 0.21648649893102492,
+      "learning_rate": 1.51578282975266e-05,
+      "loss": 0.483,
+      "step": 575
+    },
+    {
+      "epoch": 0.5233984552476147,
+      "grad_norm": 0.21837001678688414,
+      "learning_rate": 1.511273551712583e-05,
+      "loss": 0.517,
+      "step": 576
+    },
+    {
+      "epoch": 0.5243071331213085,
+      "grad_norm": 0.2193055231409076,
+      "learning_rate": 1.5067641717834358e-05,
+      "loss": 0.4995,
+      "step": 577
+    },
+    {
+      "epoch": 0.5252158109950023,
+      "grad_norm": 0.1932688376041214,
+      "learning_rate": 1.5022547307204778e-05,
+      "loss": 0.4947,
+      "step": 578
+    },
+    {
+      "epoch": 0.526124488868696,
+      "grad_norm": 0.23059329760041095,
+      "learning_rate": 1.4977452692795223e-05,
+      "loss": 0.5208,
+      "step": 579
+    },
+    {
+      "epoch": 0.5270331667423899,
+      "grad_norm": 0.2172703691219654,
+      "learning_rate": 1.4932358282165646e-05,
+      "loss": 0.5049,
+      "step": 580
+    },
+    {
+      "epoch": 0.5279418446160836,
+      "grad_norm": 0.21149601085439262,
+      "learning_rate": 1.4887264482874173e-05,
+      "loss": 0.5061,
+      "step": 581
+    },
+    {
+      "epoch": 0.5288505224897774,
+      "grad_norm": 0.22167405172460763,
+      "learning_rate": 1.4842171702473399e-05,
+      "loss": 0.5046,
+      "step": 582
+    },
+    {
+      "epoch": 0.5297592003634711,
+      "grad_norm": 0.22804024341176546,
+      "learning_rate": 1.4797080348506718e-05,
+      "loss": 0.5299,
+      "step": 583
+    },
+    {
+      "epoch": 0.5306678782371649,
+      "grad_norm": 0.2404165803826497,
+      "learning_rate": 1.4751990828504623e-05,
+      "loss": 0.5163,
+      "step": 584
+    },
+    {
+      "epoch": 0.5315765561108587,
+      "grad_norm": 0.2215909140593772,
+      "learning_rate": 1.4706903549981035e-05,
+      "loss": 0.5071,
+      "step": 585
+    },
+    {
+      "epoch": 0.5324852339845525,
+      "grad_norm": 0.2022298755605557,
+      "learning_rate": 1.466181892042962e-05,
+      "loss": 0.5233,
+      "step": 586
+    },
+    {
+      "epoch": 0.5333939118582463,
+      "grad_norm": 0.22151314608281256,
+      "learning_rate": 1.4616737347320093e-05,
+      "loss": 0.5044,
+      "step": 587
+    },
+    {
+      "epoch": 0.53430258973194,
+      "grad_norm": 0.21798627566866305,
+      "learning_rate": 1.4571659238094557e-05,
+      "loss": 0.496,
+      "step": 588
+    },
+    {
+      "epoch": 0.5352112676056338,
+      "grad_norm": 0.20560902855692786,
+      "learning_rate": 1.4526585000163802e-05,
+      "loss": 0.5013,
+      "step": 589
+    },
+    {
+      "epoch": 0.5361199454793276,
+      "grad_norm": 0.22636790116887076,
+      "learning_rate": 1.4481515040903639e-05,
+      "loss": 0.514,
+      "step": 590
+    },
+    {
+      "epoch": 0.5370286233530214,
+      "grad_norm": 0.20664895378738649,
+      "learning_rate": 1.4436449767651191e-05,
+      "loss": 0.51,
+      "step": 591
+    },
+    {
+      "epoch": 0.5379373012267151,
+      "grad_norm": 0.21391101168708013,
+      "learning_rate": 1.4391389587701245e-05,
+      "loss": 0.5091,
+      "step": 592
+    },
+    {
+      "epoch": 0.5388459791004089,
+      "grad_norm": 0.22392843939385165,
+      "learning_rate": 1.4346334908302555e-05,
+      "loss": 0.5291,
+      "step": 593
+    },
+    {
+      "epoch": 0.5397546569741026,
+      "grad_norm": 0.20828111344927408,
+      "learning_rate": 1.4301286136654146e-05,
+      "loss": 0.4944,
+      "step": 594
+    },
+    {
+      "epoch": 0.5406633348477965,
+      "grad_norm": 0.2224041235793703,
+      "learning_rate": 1.4256243679901665e-05,
+      "loss": 0.493,
+      "step": 595
+    },
+    {
+      "epoch": 0.5415720127214902,
+      "grad_norm": 0.22097320718171637,
+      "learning_rate": 1.4211207945133685e-05,
+      "loss": 0.5216,
+      "step": 596
+    },
+    {
+      "epoch": 0.542480690595184,
+      "grad_norm": 0.19560322888351336,
+      "learning_rate": 1.4166179339378022e-05,
+      "loss": 0.5001,
+      "step": 597
+    },
+    {
+      "epoch": 0.5433893684688778,
+      "grad_norm": 0.2090090684822388,
+      "learning_rate": 1.4121158269598054e-05,
+      "loss": 0.4927,
+      "step": 598
+    },
+    {
+      "epoch": 0.5442980463425715,
+      "grad_norm": 0.22417226373585894,
+      "learning_rate": 1.4076145142689065e-05,
+      "loss": 0.5283,
+      "step": 599
+    },
+    {
+      "epoch": 0.5452067242162654,
+      "grad_norm": 0.24271982418731053,
+      "learning_rate": 1.4031140365474543e-05,
+      "loss": 0.481,
+      "step": 600
+    },
+    {
+      "epoch": 0.5461154020899591,
+      "grad_norm": 0.23001454521139977,
+      "learning_rate": 1.398614434470251e-05,
+      "loss": 0.5336,
+      "step": 601
+    },
+    {
+      "epoch": 0.5470240799636529,
+      "grad_norm": 0.263483773100571,
+      "learning_rate": 1.3941157487041848e-05,
+      "loss": 0.508,
+      "step": 602
+    },
+    {
+      "epoch": 0.5479327578373466,
+      "grad_norm": 0.20548872823129907,
+      "learning_rate": 1.389618019907864e-05,
+      "loss": 0.5033,
+      "step": 603
+    },
+    {
+      "epoch": 0.5488414357110404,
+      "grad_norm": 0.23513026738046738,
+      "learning_rate": 1.385121288731246e-05,
+      "loss": 0.5101,
+      "step": 604
+    },
+    {
+      "epoch": 0.5497501135847342,
+      "grad_norm": 0.22755356478541824,
+      "learning_rate": 1.3806255958152725e-05,
+      "loss": 0.5212,
+      "step": 605
+    },
+    {
+      "epoch": 0.550658791458428,
+      "grad_norm": 0.21734920044910347,
+      "learning_rate": 1.3761309817915017e-05,
+      "loss": 0.5133,
+      "step": 606
+    },
+    {
+      "epoch": 0.5515674693321218,
+      "grad_norm": 0.22006575841027068,
+      "learning_rate": 1.3716374872817408e-05,
+      "loss": 0.4916,
+      "step": 607
+    },
+    {
+      "epoch": 0.5524761472058155,
+      "grad_norm": 0.21806084980407606,
+      "learning_rate": 1.3671451528976782e-05,
+      "loss": 0.5062,
+      "step": 608
+    },
+    {
+      "epoch": 0.5533848250795094,
+      "grad_norm": 0.26445597597573617,
+      "learning_rate": 1.3626540192405188e-05,
+      "loss": 0.5115,
+      "step": 609
+    },
+    {
+      "epoch": 0.5542935029532031,
+      "grad_norm": 0.23453239295975004,
+      "learning_rate": 1.3581641269006142e-05,
+      "loss": 0.521,
+      "step": 610
+    },
+    {
+      "epoch": 0.5552021808268969,
+      "grad_norm": 0.198955340760059,
+      "learning_rate": 1.3536755164570977e-05,
+      "loss": 0.5017,
+      "step": 611
+    },
+    {
+      "epoch": 0.5561108587005906,
+      "grad_norm": 0.2384141322061314,
+      "learning_rate": 1.3491882284775165e-05,
+      "loss": 0.5379,
+      "step": 612
+    },
+    {
+      "epoch": 0.5570195365742844,
+      "grad_norm": 0.3533164212329243,
+      "learning_rate": 1.3447023035174662e-05,
+      "loss": 0.5106,
+      "step": 613
+    },
+    {
+      "epoch": 0.5579282144479782,
+      "grad_norm": 0.23335137358636715,
+      "learning_rate": 1.340217782120222e-05,
+      "loss": 0.5029,
+      "step": 614
+    },
+    {
+      "epoch": 0.558836892321672,
+      "grad_norm": 0.2132941094220617,
+      "learning_rate": 1.3357347048163767e-05,
+      "loss": 0.5089,
+      "step": 615
+    },
+    {
+      "epoch": 0.5597455701953657,
+      "grad_norm": 0.21070445466929358,
+      "learning_rate": 1.331253112123469e-05,
+      "loss": 0.5041,
+      "step": 616
+    },
+    {
+      "epoch": 0.5606542480690595,
+      "grad_norm": 0.22376638889170286,
+      "learning_rate": 1.3267730445456208e-05,
+      "loss": 0.5256,
+      "step": 617
+    },
+    {
+      "epoch": 0.5615629259427533,
+      "grad_norm": 0.21326176673695557,
+      "learning_rate": 1.3222945425731705e-05,
+      "loss": 0.4967,
+      "step": 618
+    },
+    {
+      "epoch": 0.5624716038164471,
+      "grad_norm": 0.21837723498723205,
+      "learning_rate": 1.3178176466823057e-05,
+      "loss": 0.4915,
+      "step": 619
+    },
+    {
+      "epoch": 0.5633802816901409,
+      "grad_norm": 0.21674240459535304,
+      "learning_rate": 1.3133423973346996e-05,
+      "loss": 0.5228,
+      "step": 620
+    },
+    {
+      "epoch": 0.5642889595638346,
+      "grad_norm": 0.2089400672590786,
+      "learning_rate": 1.3088688349771425e-05,
+      "loss": 0.4983,
+      "step": 621
+    },
+    {
+      "epoch": 0.5651976374375284,
+      "grad_norm": 0.2178503682465244,
+      "learning_rate": 1.3043970000411811e-05,
+      "loss": 0.493,
+      "step": 622
+    },
+    {
+      "epoch": 0.5661063153112221,
+      "grad_norm": 0.21949719597501016,
+      "learning_rate": 1.2999269329427459e-05,
+      "loss": 0.4952,
+      "step": 623
+    },
+    {
+      "epoch": 0.567014993184916,
+      "grad_norm": 0.21843384021789625,
+      "learning_rate": 1.2954586740817927e-05,
+      "loss": 0.5005,
+      "step": 624
+    },
+    {
+      "epoch": 0.5679236710586097,
+      "grad_norm": 0.20757207890878002,
+      "learning_rate": 1.2909922638419336e-05,
+      "loss": 0.5024,
+      "step": 625
+    },
+    {
+      "epoch": 0.5688323489323035,
+      "grad_norm": 0.20851651967649487,
+      "learning_rate": 1.2865277425900725e-05,
+      "loss": 0.5044,
+      "step": 626
+    },
+    {
+      "epoch": 0.5697410268059973,
+      "grad_norm": 0.21405316788676493,
+      "learning_rate": 1.2820651506760419e-05,
+      "loss": 0.537,
+      "step": 627
+    },
+    {
+      "epoch": 0.570649704679691,
+      "grad_norm": 0.22159240306473557,
+      "learning_rate": 1.277604528432237e-05,
+      "loss": 0.5218,
+      "step": 628
+    },
+    {
+      "epoch": 0.5715583825533849,
+      "grad_norm": 0.20925418698672665,
+      "learning_rate": 1.2731459161732513e-05,
+      "loss": 0.5431,
+      "step": 629
+    },
+    {
+      "epoch": 0.5724670604270786,
+      "grad_norm": 0.21144929339922172,
+      "learning_rate": 1.2686893541955113e-05,
+      "loss": 0.5115,
+      "step": 630
+    },
+    {
+      "epoch": 0.5733757383007724,
+      "grad_norm": 0.2261368753908665,
+      "learning_rate": 1.2642348827769152e-05,
+      "loss": 0.5267,
+      "step": 631
+    },
+    {
+      "epoch": 0.5742844161744661,
+      "grad_norm": 0.20351693400728962,
+      "learning_rate": 1.2597825421764647e-05,
+      "loss": 0.5047,
+      "step": 632
+    },
+    {
+      "epoch": 0.5751930940481599,
+      "grad_norm": 0.21090532223303304,
+      "learning_rate": 1.2553323726339052e-05,
+      "loss": 0.5089,
+      "step": 633
+    },
+    {
+      "epoch": 0.5761017719218537,
+      "grad_norm": 0.21093740956542803,
+      "learning_rate": 1.2508844143693595e-05,
+      "loss": 0.5068,
+      "step": 634
+    },
+    {
+      "epoch": 0.5770104497955475,
+      "grad_norm": 0.23232720167531504,
+      "learning_rate": 1.2464387075829654e-05,
+      "loss": 0.4898,
+      "step": 635
+    },
+    {
+      "epoch": 0.5779191276692413,
+      "grad_norm": 0.20842996506303982,
+      "learning_rate": 1.2419952924545125e-05,
+      "loss": 0.5187,
+      "step": 636
+    },
+    {
+      "epoch": 0.578827805542935,
+      "grad_norm": 0.2129656762015234,
+      "learning_rate": 1.2375542091430776e-05,
+      "loss": 0.5005,
+      "step": 637
+    },
+    {
+      "epoch": 0.5797364834166288,
+      "grad_norm": 0.20955597848686514,
+      "learning_rate": 1.2331154977866638e-05,
+      "loss": 0.5039,
+      "step": 638
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 0.2153077918489364,
+      "learning_rate": 1.2286791985018356e-05,
+      "loss": 0.5143,
+      "step": 639
+    },
+    {
+      "epoch": 0.5815538391640164,
+      "grad_norm": 0.20026489084845642,
+      "learning_rate": 1.2242453513833584e-05,
+      "loss": 0.4947,
+      "step": 640
+    },
+    {
+      "epoch": 0.5824625170377101,
+      "grad_norm": 0.1983389631994418,
+      "learning_rate": 1.2198139965038356e-05,
+      "loss": 0.5286,
+      "step": 641
+    },
+    {
+      "epoch": 0.5833711949114039,
+      "grad_norm": 0.21541284455802412,
+      "learning_rate": 1.2153851739133452e-05,
+      "loss": 0.5181,
+      "step": 642
+    },
+    {
+      "epoch": 0.5842798727850976,
+      "grad_norm": 0.2286012335295052,
+      "learning_rate": 1.210958923639079e-05,
+      "loss": 0.5091,
+      "step": 643
+    },
+    {
+      "epoch": 0.5851885506587915,
+      "grad_norm": 0.21935071020420469,
+      "learning_rate": 1.20653528568498e-05,
+      "loss": 0.5168,
+      "step": 644
+    },
+    {
+      "epoch": 0.5860972285324852,
+      "grad_norm": 0.21372980356378826,
+      "learning_rate": 1.2021143000313822e-05,
+      "loss": 0.5009,
+      "step": 645
+    },
+    {
+      "epoch": 0.587005906406179,
+      "grad_norm": 0.22775805133335744,
+      "learning_rate": 1.1976960066346474e-05,
+      "loss": 0.4944,
+      "step": 646
+    },
+    {
+      "epoch": 0.5879145842798728,
+      "grad_norm": 0.24768343363930162,
+      "learning_rate": 1.1932804454268057e-05,
+      "loss": 0.5314,
+      "step": 647
+    },
+    {
+      "epoch": 0.5888232621535665,
+      "grad_norm": 0.2234310813880864,
+      "learning_rate": 1.1888676563151942e-05,
+      "loss": 0.5195,
+      "step": 648
+    },
+    {
+      "epoch": 0.5897319400272604,
+      "grad_norm": 0.2009851586835417,
+      "learning_rate": 1.184457679182096e-05,
+      "loss": 0.5122,
+      "step": 649
+    },
+    {
+      "epoch": 0.5906406179009541,
+      "grad_norm": 0.21317085471927943,
+      "learning_rate": 1.18005055388438e-05,
+      "loss": 0.4985,
+      "step": 650
+    },
+    {
+      "epoch": 0.5915492957746479,
+      "grad_norm": 0.30946744552229544,
+      "learning_rate": 1.1756463202531392e-05,
+      "loss": 0.5005,
+      "step": 651
+    },
+    {
+      "epoch": 0.5924579736483416,
+      "grad_norm": 0.23466972573336475,
+      "learning_rate": 1.1712450180933342e-05,
+      "loss": 0.5225,
+      "step": 652
+    },
+    {
+      "epoch": 0.5933666515220355,
+      "grad_norm": 0.2056155161824234,
+      "learning_rate": 1.1668466871834288e-05,
+      "loss": 0.515,
+      "step": 653
+    },
+    {
+      "epoch": 0.5942753293957292,
+      "grad_norm": 0.23023528471242866,
+      "learning_rate": 1.1624513672750358e-05,
+      "loss": 0.4999,
+      "step": 654
+    },
+    {
+      "epoch": 0.595184007269423,
+      "grad_norm": 0.21059005559472435,
+      "learning_rate": 1.1580590980925527e-05,
+      "loss": 0.4945,
+      "step": 655
+    },
+    {
+      "epoch": 0.5960926851431168,
+      "grad_norm": 0.2110617476567799,
+      "learning_rate": 1.1536699193328063e-05,
+      "loss": 0.5198,
+      "step": 656
+    },
+    {
+      "epoch": 0.5970013630168105,
+      "grad_norm": 0.22093511309521258,
+      "learning_rate": 1.1492838706646913e-05,
+      "loss": 0.5226,
+      "step": 657
+    },
+    {
+      "epoch": 0.5979100408905044,
+      "grad_norm": 0.19846566764978263,
+      "learning_rate": 1.1449009917288136e-05,
+      "loss": 0.4993,
+      "step": 658
+    },
+    {
+      "epoch": 0.5988187187641981,
+      "grad_norm": 0.20467790497594238,
+      "learning_rate": 1.1405213221371318e-05,
+      "loss": 0.4903,
+      "step": 659
+    },
+    {
+      "epoch": 0.5997273966378919,
+      "grad_norm": 0.19548151786525753,
+      "learning_rate": 1.1361449014725974e-05,
+      "loss": 0.5188,
+      "step": 660
+    },
+    {
+      "epoch": 0.6006360745115856,
+      "grad_norm": 0.1955762705848056,
+      "learning_rate": 1.1317717692888014e-05,
+      "loss": 0.4858,
+      "step": 661
+    },
+    {
+      "epoch": 0.6015447523852794,
+      "grad_norm": 0.20843634480541331,
+      "learning_rate": 1.1274019651096103e-05,
+      "loss": 0.5062,
+      "step": 662
+    },
+    {
+      "epoch": 0.6024534302589732,
+      "grad_norm": 0.22451284092117968,
+      "learning_rate": 1.123035528428815e-05,
+      "loss": 0.534,
+      "step": 663
+    },
+    {
+      "epoch": 0.603362108132667,
+      "grad_norm": 0.2136614171554175,
+      "learning_rate": 1.11867249870977e-05,
+      "loss": 0.51,
+      "step": 664
+    },
+    {
+      "epoch": 0.6042707860063607,
+      "grad_norm": 0.21850302263862784,
+      "learning_rate": 1.1143129153850385e-05,
+      "loss": 0.5172,
+      "step": 665
+    },
+    {
+      "epoch": 0.6051794638800545,
+      "grad_norm": 0.21996125799654462,
+      "learning_rate": 1.1099568178560356e-05,
+      "loss": 0.5067,
+      "step": 666
+    },
+    {
+      "epoch": 0.6060881417537483,
+      "grad_norm": 0.23093396526077578,
+      "learning_rate": 1.105604245492673e-05,
+      "loss": 0.516,
+      "step": 667
+    },
+    {
+      "epoch": 0.6069968196274421,
+      "grad_norm": 0.20379662572062934,
+      "learning_rate": 1.101255237633001e-05,
+      "loss": 0.4867,
+      "step": 668
+    },
+    {
+      "epoch": 0.6079054975011359,
+      "grad_norm": 0.19874639932072963,
+      "learning_rate": 1.0969098335828547e-05,
+      "loss": 0.4767,
+      "step": 669
+    },
+    {
+      "epoch": 0.6088141753748296,
+      "grad_norm": 0.20562885391729782,
+      "learning_rate": 1.0925680726154994e-05,
+      "loss": 0.5092,
+      "step": 670
+    },
+    {
+      "epoch": 0.6097228532485234,
+      "grad_norm": 0.18948027626777414,
+      "learning_rate": 1.0882299939712727e-05,
+      "loss": 0.5281,
+      "step": 671
+    },
+    {
+      "epoch": 0.6106315311222171,
+      "grad_norm": 0.20962122086701915,
+      "learning_rate": 1.0838956368572335e-05,
+      "loss": 0.5055,
+      "step": 672
+    },
+    {
+      "epoch": 0.611540208995911,
+      "grad_norm": 0.22852512891883703,
+      "learning_rate": 1.0795650404468047e-05,
+      "loss": 0.4874,
+      "step": 673
+    },
+    {
+      "epoch": 0.6124488868696047,
+      "grad_norm": 0.2101171091267797,
+      "learning_rate": 1.0752382438794224e-05,
+      "loss": 0.5212,
+      "step": 674
+    },
+    {
+      "epoch": 0.6133575647432985,
+      "grad_norm": 0.19332227679137917,
+      "learning_rate": 1.0709152862601775e-05,
+      "loss": 0.511,
+      "step": 675
+    },
+    {
+      "epoch": 0.6142662426169923,
+      "grad_norm": 0.20969920743518988,
+      "learning_rate": 1.0665962066594666e-05,
+      "loss": 0.4712,
+      "step": 676
+    },
+    {
+      "epoch": 0.615174920490686,
+      "grad_norm": 0.22729746345428423,
+      "learning_rate": 1.0622810441126372e-05,
+      "loss": 0.5073,
+      "step": 677
+    },
+    {
+      "epoch": 0.6160835983643799,
+      "grad_norm": 0.20566804358469726,
+      "learning_rate": 1.0579698376196339e-05,
+      "loss": 0.5066,
+      "step": 678
+    },
+    {
+      "epoch": 0.6169922762380736,
+      "grad_norm": 0.20020676818980052,
+      "learning_rate": 1.0536626261446475e-05,
+      "loss": 0.5248,
+      "step": 679
+    },
+    {
+      "epoch": 0.6179009541117674,
+      "grad_norm": 0.21155144921065883,
+      "learning_rate": 1.049359448615763e-05,
+      "loss": 0.4865,
+      "step": 680
+    },
+    {
+      "epoch": 0.6188096319854611,
+      "grad_norm": 0.23535876386613103,
+      "learning_rate": 1.0450603439246063e-05,
+      "loss": 0.5168,
+      "step": 681
+    },
+    {
+      "epoch": 0.6197183098591549,
+      "grad_norm": 0.2132538834911777,
+      "learning_rate": 1.0407653509259933e-05,
+      "loss": 0.5071,
+      "step": 682
+    },
+    {
+      "epoch": 0.6206269877328487,
+      "grad_norm": 0.21538594477453463,
+      "learning_rate": 1.036474508437579e-05,
+      "loss": 0.5149,
+      "step": 683
+    },
+    {
+      "epoch": 0.6215356656065425,
+      "grad_norm": 0.20645694082439553,
+      "learning_rate": 1.0321878552395073e-05,
+      "loss": 0.5114,
+      "step": 684
+    },
+    {
+      "epoch": 0.6224443434802363,
+      "grad_norm": 0.2281953724292079,
+      "learning_rate": 1.027905430074058e-05,
+      "loss": 0.4916,
+      "step": 685
+    },
+    {
+      "epoch": 0.62335302135393,
+      "grad_norm": 0.20946929959943567,
+      "learning_rate": 1.0236272716453012e-05,
+      "loss": 0.4813,
+      "step": 686
+    },
+    {
+      "epoch": 0.6242616992276238,
+      "grad_norm": 0.19038681295735432,
+      "learning_rate": 1.0193534186187423e-05,
+      "loss": 0.4909,
+      "step": 687
+    },
+    {
+      "epoch": 0.6251703771013176,
+      "grad_norm": 0.28388068586479065,
+      "learning_rate": 1.015083909620976e-05,
+      "loss": 0.5172,
+      "step": 688
+    },
+    {
+      "epoch": 0.6260790549750114,
+      "grad_norm": 0.19357100963981475,
+      "learning_rate": 1.0108187832393359e-05,
+      "loss": 0.4947,
+      "step": 689
+    },
+    {
+      "epoch": 0.6269877328487051,
+      "grad_norm": 0.19797701784277713,
+      "learning_rate": 1.0065580780215457e-05,
+      "loss": 0.4962,
+      "step": 690
+    },
+    {
+      "epoch": 0.6278964107223989,
+      "grad_norm": 0.1972578079022867,
+      "learning_rate": 1.0023018324753726e-05,
+      "loss": 0.4913,
+      "step": 691
+    },
+    {
+      "epoch": 0.6288050885960926,
+      "grad_norm": 0.212635477184621,
+      "learning_rate": 9.980500850682752e-06,
+      "loss": 0.5025,
+      "step": 692
+    },
+    {
+      "epoch": 0.6297137664697865,
+      "grad_norm": 0.2326728400938466,
+      "learning_rate": 9.938028742270612e-06,
+      "loss": 0.4956,
+      "step": 693
+    },
+    {
+      "epoch": 0.6306224443434802,
+      "grad_norm": 0.20880774946321953,
+      "learning_rate": 9.895602383375355e-06,
+      "loss": 0.5106,
+      "step": 694
+    },
+    {
+      "epoch": 0.631531122217174,
+      "grad_norm": 0.20984655979580405,
+      "learning_rate": 9.853222157441562e-06,
+      "loss": 0.488,
+      "step": 695
+    },
+    {
+      "epoch": 0.6324398000908678,
+      "grad_norm": 0.21557027206249252,
+      "learning_rate": 9.810888447496859e-06,
+      "loss": 0.4903,
+      "step": 696
+    },
+    {
+      "epoch": 0.6333484779645616,
+      "grad_norm": 0.262479391876784,
+      "learning_rate": 9.768601636148473e-06,
+      "loss": 0.506,
+      "step": 697
+    },
+    {
+      "epoch": 0.6342571558382554,
+      "grad_norm": 0.2168847269340925,
+      "learning_rate": 9.726362105579762e-06,
+      "loss": 0.5274,
+      "step": 698
+    },
+    {
+      "epoch": 0.6351658337119491,
+      "grad_norm": 0.2100750942914297,
+      "learning_rate": 9.684170237546783e-06,
+      "loss": 0.5095,
+      "step": 699
+    },
+    {
+      "epoch": 0.6360745115856429,
+      "grad_norm": 0.20337953864837427,
+      "learning_rate": 9.642026413374792e-06,
+      "loss": 0.5025,
+      "step": 700
+    },
+    {
+      "epoch": 0.6369831894593366,
+      "grad_norm": 0.19482375820759582,
+      "learning_rate": 9.599931013954858e-06,
+      "loss": 0.4989,
+      "step": 701
+    },
+    {
+      "epoch": 0.6378918673330305,
+      "grad_norm": 0.20122907456598915,
+      "learning_rate": 9.557884419740387e-06,
+      "loss": 0.4894,
+      "step": 702
+    },
+    {
+      "epoch": 0.6388005452067242,
+      "grad_norm": 0.21336177694342048,
+      "learning_rate": 9.515887010743677e-06,
+      "loss": 0.5042,
+      "step": 703
+    },
+    {
+      "epoch": 0.639709223080418,
+      "grad_norm": 0.21951281624554214,
+      "learning_rate": 9.473939166532506e-06,
+      "loss": 0.5129,
+      "step": 704
+    },
+    {
+      "epoch": 0.6406179009541118,
+      "grad_norm": 0.20685330838999208,
+      "learning_rate": 9.432041266226686e-06,
+      "loss": 0.4882,
+      "step": 705
+    },
+    {
+      "epoch": 0.6415265788278055,
+      "grad_norm": 0.19848178888090984,
+      "learning_rate": 9.390193688494657e-06,
+      "loss": 0.5163,
+      "step": 706
+    },
+    {
+      "epoch": 0.6424352567014994,
+      "grad_norm": 0.20764755511633418,
+      "learning_rate": 9.348396811550025e-06,
+      "loss": 0.5157,
+      "step": 707
+    },
+    {
+      "epoch": 0.6433439345751931,
+      "grad_norm": 0.20487025317138272,
+      "learning_rate": 9.306651013148185e-06,
+      "loss": 0.5115,
+      "step": 708
+    },
+    {
+      "epoch": 0.6442526124488869,
+      "grad_norm": 0.2031054214275953,
+      "learning_rate": 9.264956670582887e-06,
+      "loss": 0.5091,
+      "step": 709
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 0.18811587757096598,
+      "learning_rate": 9.223314160682821e-06,
+      "loss": 0.5009,
+      "step": 710
+    },
+    {
+      "epoch": 0.6460699681962744,
+      "grad_norm": 0.205011127608451,
+      "learning_rate": 9.181723859808225e-06,
+      "loss": 0.5251,
+      "step": 711
+    },
+    {
+      "epoch": 0.6469786460699682,
+      "grad_norm": 0.19846646697630305,
+      "learning_rate": 9.140186143847488e-06,
+      "loss": 0.5182,
+      "step": 712
+    },
+    {
+      "epoch": 0.647887323943662,
+      "grad_norm": 0.2028834532784052,
+      "learning_rate": 9.098701388213729e-06,
+      "loss": 0.4937,
+      "step": 713
+    },
+    {
+      "epoch": 0.6487960018173557,
+      "grad_norm": 0.18360750051219688,
+      "learning_rate": 9.057269967841415e-06,
+      "loss": 0.5112,
+      "step": 714
+    },
+    {
+      "epoch": 0.6497046796910495,
+      "grad_norm": 0.21458351259772931,
+      "learning_rate": 9.015892257182988e-06,
+      "loss": 0.5143,
+      "step": 715
+    },
+    {
+      "epoch": 0.6506133575647433,
+      "grad_norm": 0.2063598971840294,
+      "learning_rate": 8.974568630205462e-06,
+      "loss": 0.5196,
+      "step": 716
+    },
+    {
+      "epoch": 0.6515220354384371,
+      "grad_norm": 0.18873835348846954,
+      "learning_rate": 8.933299460387042e-06,
+      "loss": 0.5016,
+      "step": 717
+    },
+    {
+      "epoch": 0.6524307133121309,
+      "grad_norm": 0.1993299721324064,
+      "learning_rate": 8.892085120713762e-06,
+      "loss": 0.5084,
+      "step": 718
+    },
+    {
+      "epoch": 0.6533393911858246,
+      "grad_norm": 0.21144467605104506,
+      "learning_rate": 8.850925983676113e-06,
+      "loss": 0.5089,
+      "step": 719
+    },
+    {
+      "epoch": 0.6542480690595184,
+      "grad_norm": 0.20483282234928168,
+      "learning_rate": 8.809822421265665e-06,
+      "loss": 0.5014,
+      "step": 720
+    },
+    {
+      "epoch": 0.6551567469332121,
+      "grad_norm": 0.19814430014436735,
+      "learning_rate": 8.768774804971705e-06,
+      "loss": 0.5006,
+      "step": 721
+    },
+    {
+      "epoch": 0.656065424806906,
+      "grad_norm": 0.19104520609396944,
+      "learning_rate": 8.727783505777894e-06,
+      "loss": 0.5079,
+      "step": 722
+    },
+    {
+      "epoch": 0.6569741026805997,
+      "grad_norm": 0.19824918419214743,
+      "learning_rate": 8.686848894158905e-06,
+      "loss": 0.5092,
+      "step": 723
+    },
+    {
+      "epoch": 0.6578827805542935,
+      "grad_norm": 0.19627827278209117,
+      "learning_rate": 8.645971340077066e-06,
+      "loss": 0.5037,
+      "step": 724
+    },
+    {
+      "epoch": 0.6587914584279873,
+      "grad_norm": 0.21191569677398264,
+      "learning_rate": 8.605151212979033e-06,
+      "loss": 0.505,
+      "step": 725
+    },
+    {
+      "epoch": 0.659700136301681,
+      "grad_norm": 0.19339067359779558,
+      "learning_rate": 8.564388881792456e-06,
+      "loss": 0.5068,
+      "step": 726
+    },
+    {
+      "epoch": 0.6606088141753749,
+      "grad_norm": 0.20882673980251068,
+      "learning_rate": 8.523684714922608e-06,
+      "loss": 0.5209,
+      "step": 727
+    },
+    {
+      "epoch": 0.6615174920490686,
+      "grad_norm": 0.1922561192899202,
+      "learning_rate": 8.483039080249092e-06,
+      "loss": 0.4767,
+      "step": 728
+    },
+    {
+      "epoch": 0.6624261699227624,
+      "grad_norm": 0.20198892951089617,
+      "learning_rate": 8.442452345122498e-06,
+      "loss": 0.5142,
+      "step": 729
+    },
+    {
+      "epoch": 0.6633348477964561,
+      "grad_norm": 0.19893991599577202,
+      "learning_rate": 8.401924876361101e-06,
+      "loss": 0.4957,
+      "step": 730
+    },
+    {
+      "epoch": 0.6642435256701499,
+      "grad_norm": 0.18385193458028223,
+      "learning_rate": 8.361457040247518e-06,
+      "loss": 0.5059,
+      "step": 731
+    },
+    {
+      "epoch": 0.6651522035438437,
+      "grad_norm": 0.20111627364657178,
+      "learning_rate": 8.321049202525428e-06,
+      "loss": 0.5005,
+      "step": 732
+    },
+    {
+      "epoch": 0.6660608814175375,
+      "grad_norm": 0.1886288138124668,
+      "learning_rate": 8.280701728396235e-06,
+      "loss": 0.5,
+      "step": 733
+    },
+    {
+      "epoch": 0.6669695592912313,
+      "grad_norm": 0.17622798353544158,
+      "learning_rate": 8.240414982515803e-06,
+      "loss": 0.5118,
+      "step": 734
+    },
+    {
+      "epoch": 0.667878237164925,
+      "grad_norm": 0.19635146761708042,
+      "learning_rate": 8.200189328991125e-06,
+      "loss": 0.5193,
+      "step": 735
+    },
+    {
+      "epoch": 0.6687869150386188,
+      "grad_norm": 0.1897240476003535,
+      "learning_rate": 8.160025131377044e-06,
+      "loss": 0.4951,
+      "step": 736
+    },
+    {
+      "epoch": 0.6696955929123126,
+      "grad_norm": 0.19822362467607307,
+      "learning_rate": 8.119922752672994e-06,
+      "loss": 0.5045,
+      "step": 737
+    },
+    {
+      "epoch": 0.6706042707860064,
+      "grad_norm": 0.21098146480685617,
+      "learning_rate": 8.079882555319685e-06,
+      "loss": 0.5013,
+      "step": 738
+    },
+    {
+      "epoch": 0.6715129486597001,
+      "grad_norm": 0.19374590179910672,
+      "learning_rate": 8.039904901195832e-06,
+      "loss": 0.5028,
+      "step": 739
+    },
+    {
+      "epoch": 0.6724216265333939,
+      "grad_norm": 0.20027425053961329,
+      "learning_rate": 7.999990151614895e-06,
+      "loss": 0.5122,
+      "step": 740
+    },
+    {
+      "epoch": 0.6733303044070876,
+      "grad_norm": 0.20684534413735903,
+      "learning_rate": 7.960138667321822e-06,
+      "loss": 0.5074,
+      "step": 741
+    },
+    {
+      "epoch": 0.6742389822807815,
+      "grad_norm": 0.19882319560879994,
+      "learning_rate": 7.920350808489761e-06,
+      "loss": 0.4847,
+      "step": 742
+    },
+    {
+      "epoch": 0.6751476601544752,
+      "grad_norm": 0.20660789884737718,
+      "learning_rate": 7.880626934716815e-06,
+      "loss": 0.4906,
+      "step": 743
+    },
+    {
+      "epoch": 0.676056338028169,
+      "grad_norm": 0.20808779051241447,
+      "learning_rate": 7.840967405022832e-06,
+      "loss": 0.4979,
+      "step": 744
+    },
+    {
+      "epoch": 0.6769650159018628,
+      "grad_norm": 0.20415548749844553,
+      "learning_rate": 7.801372577846091e-06,
+      "loss": 0.4899,
+      "step": 745
+    },
+    {
+      "epoch": 0.6778736937755566,
+      "grad_norm": 0.22772130134969482,
+      "learning_rate": 7.76184281104011e-06,
+      "loss": 0.511,
+      "step": 746
+    },
+    {
+      "epoch": 0.6787823716492504,
+      "grad_norm": 0.19909122191248269,
+      "learning_rate": 7.72237846187039e-06,
+      "loss": 0.4991,
+      "step": 747
+    },
+    {
+      "epoch": 0.6796910495229441,
+      "grad_norm": 0.19262198987408088,
+      "learning_rate": 7.682979887011212e-06,
+      "loss": 0.4761,
+      "step": 748
+    },
+    {
+      "epoch": 0.6805997273966379,
+      "grad_norm": 0.18775529743972974,
+      "learning_rate": 7.643647442542383e-06,
+      "loss": 0.513,
+      "step": 749
+    },
+    {
+      "epoch": 0.6815084052703316,
+      "grad_norm": 0.2173785191691545,
+      "learning_rate": 7.604381483946028e-06,
+      "loss": 0.5072,
+      "step": 750
+    },
+    {
+      "epoch": 0.6824170831440255,
+      "grad_norm": 0.19147025340181428,
+      "learning_rate": 7.565182366103391e-06,
+      "loss": 0.5149,
+      "step": 751
+    },
+    {
+      "epoch": 0.6833257610177192,
+      "grad_norm": 0.21129377540626434,
+      "learning_rate": 7.526050443291621e-06,
+      "loss": 0.5057,
+      "step": 752
+    },
+    {
+      "epoch": 0.684234438891413,
+      "grad_norm": 0.20406439218144845,
+      "learning_rate": 7.486986069180552e-06,
+      "loss": 0.499,
+      "step": 753
+    },
+    {
+      "epoch": 0.6851431167651068,
+      "grad_norm": 0.18926130030422847,
+      "learning_rate": 7.447989596829518e-06,
+      "loss": 0.495,
+      "step": 754
+    },
+    {
+      "epoch": 0.6860517946388005,
+      "grad_norm": 0.20858318842417992,
+      "learning_rate": 7.409061378684187e-06,
+      "loss": 0.4938,
+      "step": 755
+    },
+    {
+      "epoch": 0.6869604725124944,
+      "grad_norm": 0.2131058546708383,
+      "learning_rate": 7.370201766573325e-06,
+      "loss": 0.4913,
+      "step": 756
+    },
+    {
+      "epoch": 0.6878691503861881,
+      "grad_norm": 0.20230975445307423,
+      "learning_rate": 7.331411111705673e-06,
+      "loss": 0.4876,
+      "step": 757
+    },
+    {
+      "epoch": 0.6887778282598819,
+      "grad_norm": 0.2053682321742907,
+      "learning_rate": 7.2926897646667185e-06,
+      "loss": 0.5141,
+      "step": 758
+    },
+    {
+      "epoch": 0.6896865061335756,
+      "grad_norm": 0.1954233893947322,
+      "learning_rate": 7.254038075415573e-06,
+      "loss": 0.5183,
+      "step": 759
+    },
+    {
+      "epoch": 0.6905951840072694,
+      "grad_norm": 0.18580034606326304,
+      "learning_rate": 7.215456393281777e-06,
+      "loss": 0.4906,
+      "step": 760
+    },
+    {
+      "epoch": 0.6915038618809632,
+      "grad_norm": 0.19465847889916651,
+      "learning_rate": 7.176945066962152e-06,
+      "loss": 0.5151,
+      "step": 761
+    },
+    {
+      "epoch": 0.692412539754657,
+      "grad_norm": 0.19359527460822998,
+      "learning_rate": 7.138504444517669e-06,
+      "loss": 0.5101,
+      "step": 762
+    },
+    {
+      "epoch": 0.6933212176283507,
+      "grad_norm": 0.2015981354614617,
+      "learning_rate": 7.100134873370262e-06,
+      "loss": 0.5119,
+      "step": 763
+    },
+    {
+      "epoch": 0.6942298955020445,
+      "grad_norm": 0.19680747451784078,
+      "learning_rate": 7.061836700299735e-06,
+      "loss": 0.4846,
+      "step": 764
+    },
+    {
+      "epoch": 0.6951385733757383,
+      "grad_norm": 0.20697700768347255,
+      "learning_rate": 7.023610271440583e-06,
+      "loss": 0.5047,
+      "step": 765
+    },
+    {
+      "epoch": 0.6960472512494321,
+      "grad_norm": 0.19041125662893202,
+      "learning_rate": 6.985455932278904e-06,
+      "loss": 0.5041,
+      "step": 766
+    },
+    {
+      "epoch": 0.6969559291231259,
+      "grad_norm": 0.1903949228408043,
+      "learning_rate": 6.947374027649247e-06,
+      "loss": 0.5234,
+      "step": 767
+    },
+    {
+      "epoch": 0.6978646069968196,
+      "grad_norm": 0.1806764266668593,
+      "learning_rate": 6.909364901731498e-06,
+      "loss": 0.4892,
+      "step": 768
+    },
+    {
+      "epoch": 0.6987732848705134,
+      "grad_norm": 0.18271924249493374,
+      "learning_rate": 6.871428898047793e-06,
+      "loss": 0.4893,
+      "step": 769
+    },
+    {
+      "epoch": 0.6996819627442071,
+      "grad_norm": 0.19917497101420706,
+      "learning_rate": 6.833566359459399e-06,
+      "loss": 0.5105,
+      "step": 770
+    },
+    {
+      "epoch": 0.700590640617901,
+      "grad_norm": 0.1863020335714146,
+      "learning_rate": 6.795777628163599e-06,
+      "loss": 0.5056,
+      "step": 771
+    },
+    {
+      "epoch": 0.7014993184915947,
+      "grad_norm": 0.18704918812911453,
+      "learning_rate": 6.758063045690618e-06,
+      "loss": 0.5128,
+      "step": 772
+    },
+    {
+      "epoch": 0.7024079963652885,
+      "grad_norm": 0.19036154965202196,
+      "learning_rate": 6.720422952900545e-06,
+      "loss": 0.5109,
+      "step": 773
+    },
+    {
+      "epoch": 0.7033166742389823,
+      "grad_norm": 0.19259035264174634,
+      "learning_rate": 6.6828576899802265e-06,
+      "loss": 0.4928,
+      "step": 774
+    },
+    {
+      "epoch": 0.704225352112676,
+      "grad_norm": 0.21312159007001835,
+      "learning_rate": 6.645367596440202e-06,
+      "loss": 0.5158,
+      "step": 775
+    },
+    {
+      "epoch": 0.7051340299863699,
+      "grad_norm": 0.20251937662605798,
+      "learning_rate": 6.607953011111655e-06,
+      "loss": 0.4764,
+      "step": 776
+    },
+    {
+      "epoch": 0.7060427078600636,
+      "grad_norm": 0.18240217297545883,
+      "learning_rate": 6.570614272143327e-06,
+      "loss": 0.4903,
+      "step": 777
+    },
+    {
+      "epoch": 0.7069513857337574,
+      "grad_norm": 0.20130547911212013,
+      "learning_rate": 6.533351716998466e-06,
+      "loss": 0.4906,
+      "step": 778
+    },
+    {
+      "epoch": 0.7078600636074511,
+      "grad_norm": 0.1959544034246125,
+      "learning_rate": 6.496165682451772e-06,
+      "loss": 0.5112,
+      "step": 779
+    },
+    {
+      "epoch": 0.7087687414811449,
+      "grad_norm": 0.19921216780711878,
+      "learning_rate": 6.459056504586382e-06,
+      "loss": 0.523,
+      "step": 780
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 0.19936125578543795,
+      "learning_rate": 6.4220245187907915e-06,
+      "loss": 0.5229,
+      "step": 781
+    },
+    {
+      "epoch": 0.7105860972285325,
+      "grad_norm": 0.20670310351708626,
+      "learning_rate": 6.3850700597558465e-06,
+      "loss": 0.499,
+      "step": 782
+    },
+    {
+      "epoch": 0.7114947751022263,
+      "grad_norm": 0.19209470383493454,
+      "learning_rate": 6.348193461471723e-06,
+      "loss": 0.4991,
+      "step": 783
+    },
+    {
+      "epoch": 0.71240345297592,
+      "grad_norm": 0.18972382968362497,
+      "learning_rate": 6.3113950572248995e-06,
+      "loss": 0.5097,
+      "step": 784
+    },
+    {
+      "epoch": 0.7133121308496138,
+      "grad_norm": 0.20452856428727872,
+      "learning_rate": 6.274675179595138e-06,
+      "loss": 0.5056,
+      "step": 785
+    },
+    {
+      "epoch": 0.7142208087233076,
+      "grad_norm": 0.19987737706875874,
+      "learning_rate": 6.238034160452486e-06,
+      "loss": 0.5106,
+      "step": 786
+    },
+    {
+      "epoch": 0.7151294865970014,
+      "grad_norm": 0.20681260557647593,
+      "learning_rate": 6.2014723309542885e-06,
+      "loss": 0.4914,
+      "step": 787
+    },
+    {
+      "epoch": 0.7160381644706951,
+      "grad_norm": 0.18866635094855325,
+      "learning_rate": 6.164990021542173e-06,
+      "loss": 0.4946,
+      "step": 788
+    },
+    {
+      "epoch": 0.7169468423443889,
+      "grad_norm": 0.20387756471884397,
+      "learning_rate": 6.128587561939069e-06,
+      "loss": 0.5067,
+      "step": 789
+    },
+    {
+      "epoch": 0.7178555202180827,
+      "grad_norm": 0.2071972801022147,
+      "learning_rate": 6.092265281146247e-06,
+      "loss": 0.5013,
+      "step": 790
+    },
+    {
+      "epoch": 0.7187641980917765,
+      "grad_norm": 0.1949470338273329,
+      "learning_rate": 6.05602350744033e-06,
+      "loss": 0.4913,
+      "step": 791
+    },
+    {
+      "epoch": 0.7196728759654702,
+      "grad_norm": 0.18810199549074258,
+      "learning_rate": 6.0198625683703155e-06,
+      "loss": 0.5105,
+      "step": 792
+    },
+    {
+      "epoch": 0.720581553839164,
+      "grad_norm": 0.19860832355578975,
+      "learning_rate": 5.983782790754624e-06,
+      "loss": 0.5136,
+      "step": 793
+    },
+    {
+      "epoch": 0.7214902317128578,
+      "grad_norm": 0.20946392891516918,
+      "learning_rate": 5.94778450067817e-06,
+      "loss": 0.5229,
+      "step": 794
+    },
+    {
+      "epoch": 0.7223989095865516,
+      "grad_norm": 0.19397171828229967,
+      "learning_rate": 5.911868023489367e-06,
+      "loss": 0.5027,
+      "step": 795
+    },
+    {
+      "epoch": 0.7233075874602454,
+      "grad_norm": 0.20693216539779685,
+      "learning_rate": 5.8760336837972355e-06,
+      "loss": 0.4838,
+      "step": 796
+    },
+    {
+      "epoch": 0.7242162653339391,
+      "grad_norm": 0.20734602876695452,
+      "learning_rate": 5.840281805468427e-06,
+      "loss": 0.4802,
+      "step": 797
+    },
+    {
+      "epoch": 0.7251249432076329,
+      "grad_norm": 0.1899623876568743,
+      "learning_rate": 5.804612711624336e-06,
+      "loss": 0.5277,
+      "step": 798
+    },
+    {
+      "epoch": 0.7260336210813266,
+      "grad_norm": 0.17933405146834225,
+      "learning_rate": 5.769026724638145e-06,
+      "loss": 0.5018,
+      "step": 799
+    },
+    {
+      "epoch": 0.7269422989550205,
+      "grad_norm": 0.19234978551841816,
+      "learning_rate": 5.733524166131927e-06,
+      "loss": 0.4977,
+      "step": 800
+    },
+    {
+      "epoch": 0.7278509768287142,
+      "grad_norm": 0.19968365635750715,
+      "learning_rate": 5.6981053569737525e-06,
+      "loss": 0.51,
+      "step": 801
+    },
+    {
+      "epoch": 0.728759654702408,
+      "grad_norm": 0.19239597936997574,
+      "learning_rate": 5.662770617274754e-06,
+      "loss": 0.5171,
+      "step": 802
+    },
+    {
+      "epoch": 0.7296683325761018,
+      "grad_norm": 0.21370083537965098,
+      "learning_rate": 5.627520266386275e-06,
+      "loss": 0.5181,
+      "step": 803
+    },
+    {
+      "epoch": 0.7305770104497955,
+      "grad_norm": 0.18966960367117558,
+      "learning_rate": 5.592354622896944e-06,
+      "loss": 0.4914,
+      "step": 804
+    },
+    {
+      "epoch": 0.7314856883234894,
+      "grad_norm": 0.33869822983935133,
+      "learning_rate": 5.557274004629831e-06,
+      "loss": 0.52,
+      "step": 805
+    },
+    {
+      "epoch": 0.7323943661971831,
+      "grad_norm": 0.2394910020691673,
+      "learning_rate": 5.522278728639544e-06,
+      "loss": 0.5106,
+      "step": 806
+    },
+    {
+      "epoch": 0.7333030440708769,
+      "grad_norm": 0.19290629969638587,
+      "learning_rate": 5.487369111209378e-06,
+      "loss": 0.5161,
+      "step": 807
+    },
+    {
+      "epoch": 0.7342117219445706,
+      "grad_norm": 0.19861433034388595,
+      "learning_rate": 5.4525454678484625e-06,
+      "loss": 0.508,
+      "step": 808
+    },
+    {
+      "epoch": 0.7351203998182644,
+      "grad_norm": 0.18958071403243007,
+      "learning_rate": 5.417808113288907e-06,
+      "loss": 0.4938,
+      "step": 809
+    },
+    {
+      "epoch": 0.7360290776919582,
+      "grad_norm": 0.2597240847120721,
+      "learning_rate": 5.3831573614829445e-06,
+      "loss": 0.507,
+      "step": 810
+    },
+    {
+      "epoch": 0.736937755565652,
+      "grad_norm": 0.19300705218252923,
+      "learning_rate": 5.348593525600093e-06,
+      "loss": 0.5026,
+      "step": 811
+    },
+    {
+      "epoch": 0.7378464334393458,
+      "grad_norm": 0.2017457758270549,
+      "learning_rate": 5.31411691802436e-06,
+      "loss": 0.5235,
+      "step": 812
+    },
+    {
+      "epoch": 0.7387551113130395,
+      "grad_norm": 0.18594427403779049,
+      "learning_rate": 5.279727850351373e-06,
+      "loss": 0.4953,
+      "step": 813
+    },
+    {
+      "epoch": 0.7396637891867333,
+      "grad_norm": 0.18821737436264005,
+      "learning_rate": 5.245426633385583e-06,
+      "loss": 0.5017,
+      "step": 814
+    },
+    {
+      "epoch": 0.7405724670604271,
+      "grad_norm": 0.18278935814928868,
+      "learning_rate": 5.21121357713747e-06,
+      "loss": 0.5227,
+      "step": 815
+    },
+    {
+      "epoch": 0.7414811449341209,
+      "grad_norm": 0.2467856622630847,
+      "learning_rate": 5.177088990820725e-06,
+      "loss": 0.5161,
+      "step": 816
+    },
+    {
+      "epoch": 0.7423898228078146,
+      "grad_norm": 0.19323901146064826,
+      "learning_rate": 5.1430531828494476e-06,
+      "loss": 0.5253,
+      "step": 817
+    },
+    {
+      "epoch": 0.7432985006815084,
+      "grad_norm": 0.1882470146628379,
+      "learning_rate": 5.109106460835373e-06,
+      "loss": 0.4987,
+      "step": 818
+    },
+    {
+      "epoch": 0.7442071785552021,
+      "grad_norm": 0.19056117982874113,
+      "learning_rate": 5.075249131585095e-06,
+      "loss": 0.5047,
+      "step": 819
+    },
+    {
+      "epoch": 0.745115856428896,
+      "grad_norm": 0.1830098479521597,
+      "learning_rate": 5.04148150109728e-06,
+      "loss": 0.4853,
+      "step": 820
+    },
+    {
+      "epoch": 0.7460245343025897,
+      "grad_norm": 0.22439208981690162,
+      "learning_rate": 5.0078038745599e-06,
+      "loss": 0.5064,
+      "step": 821
+    },
+    {
+      "epoch": 0.7469332121762835,
+      "grad_norm": 0.19494742516744892,
+      "learning_rate": 4.9742165563474945e-06,
+      "loss": 0.4938,
+      "step": 822
+    },
+    {
+      "epoch": 0.7478418900499773,
+      "grad_norm": 0.18946082176821574,
+      "learning_rate": 4.940719850018406e-06,
+      "loss": 0.4969,
+      "step": 823
+    },
+    {
+      "epoch": 0.748750567923671,
+      "grad_norm": 0.19406823767651848,
+      "learning_rate": 4.907314058312029e-06,
+      "loss": 0.5002,
+      "step": 824
+    },
+    {
+      "epoch": 0.7496592457973649,
+      "grad_norm": 0.1878236850523878,
+      "learning_rate": 4.8739994831460795e-06,
+      "loss": 0.5135,
+      "step": 825
+    },
+    {
+      "epoch": 0.7505679236710586,
+      "grad_norm": 0.18632298196275213,
+      "learning_rate": 4.840776425613887e-06,
+      "loss": 0.5021,
+      "step": 826
+    },
+    {
+      "epoch": 0.7514766015447524,
+      "grad_norm": 0.18739694082021402,
+      "learning_rate": 4.807645185981629e-06,
+      "loss": 0.5009,
+      "step": 827
+    },
+    {
+      "epoch": 0.7523852794184461,
+      "grad_norm": 0.19406009330598228,
+      "learning_rate": 4.774606063685671e-06,
+      "loss": 0.4979,
+      "step": 828
+    },
+    {
+      "epoch": 0.7532939572921399,
+      "grad_norm": 0.2033133910970134,
+      "learning_rate": 4.74165935732981e-06,
+      "loss": 0.5149,
+      "step": 829
+    },
+    {
+      "epoch": 0.7542026351658337,
+      "grad_norm": 0.18694027292685755,
+      "learning_rate": 4.708805364682613e-06,
+      "loss": 0.487,
+      "step": 830
+    },
+    {
+      "epoch": 0.7551113130395275,
+      "grad_norm": 0.19488605237165357,
+      "learning_rate": 4.676044382674702e-06,
+      "loss": 0.5134,
+      "step": 831
+    },
+    {
+      "epoch": 0.7560199909132213,
+      "grad_norm": 0.18325539474678315,
+      "learning_rate": 4.64337670739608e-06,
+      "loss": 0.5067,
+      "step": 832
+    },
+    {
+      "epoch": 0.756928668786915,
+      "grad_norm": 0.19497076360522683,
+      "learning_rate": 4.610802634093466e-06,
+      "loss": 0.5059,
+      "step": 833
+    },
+    {
+      "epoch": 0.7578373466606089,
+      "grad_norm": 0.17460360852298493,
+      "learning_rate": 4.578322457167594e-06,
+      "loss": 0.498,
+      "step": 834
+    },
+    {
+      "epoch": 0.7587460245343026,
+      "grad_norm": 0.1872301992724812,
+      "learning_rate": 4.5459364701706e-06,
+      "loss": 0.5074,
+      "step": 835
+    },
+    {
+      "epoch": 0.7596547024079964,
+      "grad_norm": 0.1862065608820259,
+      "learning_rate": 4.513644965803316e-06,
+      "loss": 0.5076,
+      "step": 836
+    },
+    {
+      "epoch": 0.7605633802816901,
+      "grad_norm": 0.18412311731882353,
+      "learning_rate": 4.481448235912671e-06,
+      "loss": 0.4986,
+      "step": 837
+    },
+    {
+      "epoch": 0.7614720581553839,
+      "grad_norm": 0.18372335213077792,
+      "learning_rate": 4.449346571489024e-06,
+      "loss": 0.5062,
+      "step": 838
+    },
+    {
+      "epoch": 0.7623807360290777,
+      "grad_norm": 0.18912332542943594,
+      "learning_rate": 4.417340262663533e-06,
+      "loss": 0.5299,
+      "step": 839
+    },
+    {
+      "epoch": 0.7632894139027715,
+      "grad_norm": 0.1869943803405463,
+      "learning_rate": 4.385429598705566e-06,
+      "loss": 0.502,
+      "step": 840
+    },
+    {
+      "epoch": 0.7641980917764652,
+      "grad_norm": 0.17811159329125326,
+      "learning_rate": 4.353614868020051e-06,
+      "loss": 0.4949,
+      "step": 841
+    },
+    {
+      "epoch": 0.765106769650159,
+      "grad_norm": 0.17061573326856555,
+      "learning_rate": 4.3218963581448805e-06,
+      "loss": 0.506,
+      "step": 842
+    },
+    {
+      "epoch": 0.7660154475238528,
+      "grad_norm": 0.19053231471949375,
+      "learning_rate": 4.290274355748312e-06,
+      "loss": 0.495,
+      "step": 843
+    },
+    {
+      "epoch": 0.7669241253975466,
+      "grad_norm": 0.20701343568348013,
+      "learning_rate": 4.258749146626391e-06,
+      "loss": 0.4916,
+      "step": 844
+    },
+    {
+      "epoch": 0.7678328032712404,
+      "grad_norm": 0.19297431569075968,
+      "learning_rate": 4.227321015700348e-06,
+      "loss": 0.5079,
+      "step": 845
+    },
+    {
+      "epoch": 0.7687414811449341,
+      "grad_norm": 0.17863142407970647,
+      "learning_rate": 4.195990247014025e-06,
+      "loss": 0.4922,
+      "step": 846
+    },
+    {
+      "epoch": 0.7696501590186279,
+      "grad_norm": 0.17444816686104248,
+      "learning_rate": 4.164757123731332e-06,
+      "loss": 0.5111,
+      "step": 847
+    },
+    {
+      "epoch": 0.7705588368923216,
+      "grad_norm": 0.18582842524471507,
+      "learning_rate": 4.133621928133666e-06,
+      "loss": 0.504,
+      "step": 848
+    },
+    {
+      "epoch": 0.7714675147660155,
+      "grad_norm": 0.19319239687061338,
+      "learning_rate": 4.102584941617356e-06,
+      "loss": 0.5138,
+      "step": 849
+    },
+    {
+      "epoch": 0.7723761926397092,
+      "grad_norm": 0.18983118802711624,
+      "learning_rate": 4.07164644469113e-06,
+      "loss": 0.5116,
+      "step": 850
+    },
+    {
+      "epoch": 0.773284870513403,
+      "grad_norm": 0.1822008351253182,
+      "learning_rate": 4.04080671697359e-06,
+      "loss": 0.5241,
+      "step": 851
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.19588345262435536,
+      "learning_rate": 4.010066037190664e-06,
+      "loss": 0.4892,
+      "step": 852
+    },
+    {
+      "epoch": 0.7751022262607905,
+      "grad_norm": 0.18127769473449273,
+      "learning_rate": 3.97942468317309e-06,
+      "loss": 0.4979,
+      "step": 853
+    },
+    {
+      "epoch": 0.7760109041344844,
+      "grad_norm": 0.17747552772155908,
+      "learning_rate": 3.948882931853924e-06,
+      "loss": 0.5038,
+      "step": 854
+    },
+    {
+      "epoch": 0.7769195820081781,
+      "grad_norm": 0.17474735068767536,
+      "learning_rate": 3.918441059266026e-06,
+      "loss": 0.5055,
+      "step": 855
+    },
+    {
+      "epoch": 0.7778282598818719,
+      "grad_norm": 0.18110316556024372,
+      "learning_rate": 3.888099340539548e-06,
+      "loss": 0.5142,
+      "step": 856
+    },
+    {
+      "epoch": 0.7787369377555656,
+      "grad_norm": 0.17381313511095411,
+      "learning_rate": 3.857858049899469e-06,
+      "loss": 0.5008,
+      "step": 857
+    },
+    {
+      "epoch": 0.7796456156292594,
+      "grad_norm": 0.2058649846312629,
+      "learning_rate": 3.827717460663122e-06,
+      "loss": 0.5171,
+      "step": 858
+    },
+    {
+      "epoch": 0.7805542935029532,
+      "grad_norm": 0.1994228195418579,
+      "learning_rate": 3.7976778452376966e-06,
+      "loss": 0.5121,
+      "step": 859
+    },
+    {
+      "epoch": 0.781462971376647,
+      "grad_norm": 0.19106219581068892,
+      "learning_rate": 3.7677394751177957e-06,
+      "loss": 0.5014,
+      "step": 860
+    },
+    {
+      "epoch": 0.7823716492503408,
+      "grad_norm": 0.18060774785185574,
+      "learning_rate": 3.7379026208829865e-06,
+      "loss": 0.4961,
+      "step": 861
+    },
+    {
+      "epoch": 0.7832803271240345,
+      "grad_norm": 0.19089962854174491,
+      "learning_rate": 3.708167552195346e-06,
+      "loss": 0.5083,
+      "step": 862
+    },
+    {
+      "epoch": 0.7841890049977283,
+      "grad_norm": 0.24135358768565288,
+      "learning_rate": 3.67853453779702e-06,
+      "loss": 0.5083,
+      "step": 863
+    },
+    {
+      "epoch": 0.7850976828714221,
+      "grad_norm": 0.1971222558984969,
+      "learning_rate": 3.6490038455077933e-06,
+      "loss": 0.4935,
+      "step": 864
+    },
+    {
+      "epoch": 0.7860063607451159,
+      "grad_norm": 0.19382482178376545,
+      "learning_rate": 3.619575742222695e-06,
+      "loss": 0.4829,
+      "step": 865
+    },
+    {
+      "epoch": 0.7869150386188096,
+      "grad_norm": 0.19118227592361328,
+      "learning_rate": 3.5902504939095444e-06,
+      "loss": 0.4944,
+      "step": 866
+    },
+    {
+      "epoch": 0.7878237164925034,
+      "grad_norm": 0.18091993720611124,
+      "learning_rate": 3.561028365606588e-06,
+      "loss": 0.5216,
+      "step": 867
+    },
+    {
+      "epoch": 0.7887323943661971,
+      "grad_norm": 0.175795228137756,
+      "learning_rate": 3.5319096214200674e-06,
+      "loss": 0.4925,
+      "step": 868
+    },
+    {
+      "epoch": 0.789641072239891,
+      "grad_norm": 0.18902176692875,
+      "learning_rate": 3.5028945245218705e-06,
+      "loss": 0.5055,
+      "step": 869
+    },
+    {
+      "epoch": 0.7905497501135847,
+      "grad_norm": 0.18848539499883823,
+      "learning_rate": 3.473983337147118e-06,
+      "loss": 0.5096,
+      "step": 870
+    },
+    {
+      "epoch": 0.7914584279872785,
+      "grad_norm": 0.17412933715596032,
+      "learning_rate": 3.445176320591806e-06,
+      "loss": 0.4939,
+      "step": 871
+    },
+    {
+      "epoch": 0.7923671058609723,
+      "grad_norm": 0.17984972956365003,
+      "learning_rate": 3.4164737352104635e-06,
+      "loss": 0.5103,
+      "step": 872
+    },
+    {
+      "epoch": 0.793275783734666,
+      "grad_norm": 0.17454380790489255,
+      "learning_rate": 3.3878758404137627e-06,
+      "loss": 0.5069,
+      "step": 873
+    },
+    {
+      "epoch": 0.7941844616083599,
+      "grad_norm": 0.17501932127393316,
+      "learning_rate": 3.359382894666213e-06,
+      "loss": 0.4994,
+      "step": 874
+    },
+    {
+      "epoch": 0.7950931394820536,
+      "grad_norm": 0.19096499205483844,
+      "learning_rate": 3.3309951554837882e-06,
+      "loss": 0.48,
+      "step": 875
+    },
+    {
+      "epoch": 0.7960018173557474,
+      "grad_norm": 0.18864955376852108,
+      "learning_rate": 3.3027128794316353e-06,
+      "loss": 0.509,
+      "step": 876
+    },
+    {
+      "epoch": 0.7969104952294411,
+      "grad_norm": 0.18909849882862032,
+      "learning_rate": 3.274536322121723e-06,
+      "loss": 0.525,
+      "step": 877
+    },
+    {
+      "epoch": 0.797819173103135,
+      "grad_norm": 0.18298262301252236,
+      "learning_rate": 3.2464657382105484e-06,
+      "loss": 0.508,
+      "step": 878
+    },
+    {
+      "epoch": 0.7987278509768287,
+      "grad_norm": 0.19031703336610836,
+      "learning_rate": 3.2185013813968388e-06,
+      "loss": 0.4913,
+      "step": 879
+    },
+    {
+      "epoch": 0.7996365288505225,
+      "grad_norm": 0.17991906333216653,
+      "learning_rate": 3.1906435044192545e-06,
+      "loss": 0.4957,
+      "step": 880
+    },
+    {
+      "epoch": 0.8005452067242163,
+      "grad_norm": 0.18203061589219746,
+      "learning_rate": 3.162892359054098e-06,
+      "loss": 0.5276,
+      "step": 881
+    },
+    {
+      "epoch": 0.80145388459791,
+      "grad_norm": 0.179165724080468,
+      "learning_rate": 3.1352481961130382e-06,
+      "loss": 0.4954,
+      "step": 882
+    },
+    {
+      "epoch": 0.8023625624716039,
+      "grad_norm": 0.19297597464700594,
+      "learning_rate": 3.107711265440866e-06,
+      "loss": 0.489,
+      "step": 883
+    },
+    {
+      "epoch": 0.8032712403452976,
+      "grad_norm": 0.17949577061893357,
+      "learning_rate": 3.080281815913203e-06,
+      "loss": 0.5056,
+      "step": 884
+    },
+    {
+      "epoch": 0.8041799182189914,
+      "grad_norm": 0.18672578901663653,
+      "learning_rate": 3.0529600954342717e-06,
+      "loss": 0.5215,
+      "step": 885
+    },
+    {
+      "epoch": 0.8050885960926851,
+      "grad_norm": 0.18219077965912273,
+      "learning_rate": 3.02574635093466e-06,
+      "loss": 0.4943,
+      "step": 886
+    },
+    {
+      "epoch": 0.8059972739663789,
+      "grad_norm": 0.1849602022616444,
+      "learning_rate": 2.998640828369079e-06,
+      "loss": 0.4999,
+      "step": 887
+    },
+    {
+      "epoch": 0.8069059518400727,
+      "grad_norm": 0.18588003223304814,
+      "learning_rate": 2.9716437727141348e-06,
+      "loss": 0.5125,
+      "step": 888
+    },
+    {
+      "epoch": 0.8078146297137665,
+      "grad_norm": 0.1828707964089993,
+      "learning_rate": 2.9447554279661236e-06,
+      "loss": 0.5237,
+      "step": 889
+    },
+    {
+      "epoch": 0.8087233075874602,
+      "grad_norm": 0.1834448257924201,
+      "learning_rate": 2.9179760371388347e-06,
+      "loss": 0.5028,
+      "step": 890
+    },
+    {
+      "epoch": 0.809631985461154,
+      "grad_norm": 0.1822724655343291,
+      "learning_rate": 2.8913058422613363e-06,
+      "loss": 0.4882,
+      "step": 891
+    },
+    {
+      "epoch": 0.8105406633348478,
+      "grad_norm": 0.22324245823221206,
+      "learning_rate": 2.86474508437579e-06,
+      "loss": 0.495,
+      "step": 892
+    },
+    {
+      "epoch": 0.8114493412085416,
+      "grad_norm": 0.1785989609888126,
+      "learning_rate": 2.8382940035352913e-06,
+      "loss": 0.5016,
+      "step": 893
+    },
+    {
+      "epoch": 0.8123580190822354,
+      "grad_norm": 0.182834626624704,
+      "learning_rate": 2.811952838801685e-06,
+      "loss": 0.4924,
+      "step": 894
+    },
+    {
+      "epoch": 0.8132666969559291,
+      "grad_norm": 0.1703850230922861,
+      "learning_rate": 2.785721828243399e-06,
+      "loss": 0.4995,
+      "step": 895
+    },
+    {
+      "epoch": 0.8141753748296229,
+      "grad_norm": 0.17624457325668033,
+      "learning_rate": 2.7596012089333015e-06,
+      "loss": 0.4966,
+      "step": 896
+    },
+    {
+      "epoch": 0.8150840527033166,
+      "grad_norm": 0.21894574439915565,
+      "learning_rate": 2.733591216946568e-06,
+      "loss": 0.4994,
+      "step": 897
+    },
+    {
+      "epoch": 0.8159927305770105,
+      "grad_norm": 0.1813886036339837,
+      "learning_rate": 2.707692087358521e-06,
+      "loss": 0.5029,
+      "step": 898
+    },
+    {
+      "epoch": 0.8169014084507042,
+      "grad_norm": 0.2152869543169518,
+      "learning_rate": 2.681904054242537e-06,
+      "loss": 0.4945,
+      "step": 899
+    },
+    {
+      "epoch": 0.817810086324398,
+      "grad_norm": 0.18222344404245855,
+      "learning_rate": 2.6562273506679007e-06,
+      "loss": 0.4936,
+      "step": 900
+    },
+    {
+      "epoch": 0.8187187641980918,
+      "grad_norm": 0.17607757385435371,
+      "learning_rate": 2.6306622086977288e-06,
+      "loss": 0.5062,
+      "step": 901
+    },
+    {
+      "epoch": 0.8196274420717855,
+      "grad_norm": 0.19606345598832053,
+      "learning_rate": 2.6052088593868435e-06,
+      "loss": 0.5172,
+      "step": 902
+    },
+    {
+      "epoch": 0.8205361199454794,
+      "grad_norm": 0.19039708237380526,
+      "learning_rate": 2.5798675327796994e-06,
+      "loss": 0.5086,
+      "step": 903
+    },
+    {
+      "epoch": 0.8214447978191731,
+      "grad_norm": 0.18246641597501811,
+      "learning_rate": 2.554638457908314e-06,
+      "loss": 0.5082,
+      "step": 904
+    },
+    {
+      "epoch": 0.8223534756928669,
+      "grad_norm": 0.18183943600553937,
+      "learning_rate": 2.5295218627901695e-06,
+      "loss": 0.5109,
+      "step": 905
+    },
+    {
+      "epoch": 0.8232621535665606,
+      "grad_norm": 0.17028064701481918,
+      "learning_rate": 2.5045179744261864e-06,
+      "loss": 0.504,
+      "step": 906
+    },
+    {
+      "epoch": 0.8241708314402544,
+      "grad_norm": 0.18673438138373802,
+      "learning_rate": 2.4796270187986393e-06,
+      "loss": 0.5253,
+      "step": 907
+    },
+    {
+      "epoch": 0.8250795093139482,
+      "grad_norm": 0.17265401886870885,
+      "learning_rate": 2.454849220869142e-06,
+      "loss": 0.4996,
+      "step": 908
+    },
+    {
+      "epoch": 0.825988187187642,
+      "grad_norm": 0.17469166767696584,
+      "learning_rate": 2.430184804576594e-06,
+      "loss": 0.5119,
+      "step": 909
+    },
+    {
+      "epoch": 0.8268968650613358,
+      "grad_norm": 0.18899123379640204,
+      "learning_rate": 2.405633992835164e-06,
+      "loss": 0.4904,
+      "step": 910
+    },
+    {
+      "epoch": 0.8278055429350295,
+      "grad_norm": 0.18294001642802674,
+      "learning_rate": 2.38119700753228e-06,
+      "loss": 0.5025,
+      "step": 911
+    },
+    {
+      "epoch": 0.8287142208087233,
+      "grad_norm": 0.17547004300065802,
+      "learning_rate": 2.3568740695266246e-06,
+      "loss": 0.5094,
+      "step": 912
+    },
+    {
+      "epoch": 0.8296228986824171,
+      "grad_norm": 0.1752961322019241,
+      "learning_rate": 2.332665398646122e-06,
+      "loss": 0.5171,
+      "step": 913
+    },
+    {
+      "epoch": 0.8305315765561109,
+      "grad_norm": 0.1869406585448319,
+      "learning_rate": 2.308571213685967e-06,
+      "loss": 0.5053,
+      "step": 914
+    },
+    {
+      "epoch": 0.8314402544298046,
+      "grad_norm": 0.17875534656100106,
+      "learning_rate": 2.2845917324066522e-06,
+      "loss": 0.4994,
+      "step": 915
+    },
+    {
+      "epoch": 0.8323489323034984,
+      "grad_norm": 0.18840384284937645,
+      "learning_rate": 2.260727171531982e-06,
+      "loss": 0.5187,
+      "step": 916
+    },
+    {
+      "epoch": 0.8332576101771921,
+      "grad_norm": 0.17687582668078677,
+      "learning_rate": 2.2369777467471198e-06,
+      "loss": 0.4915,
+      "step": 917
+    },
+    {
+      "epoch": 0.834166288050886,
+      "grad_norm": 0.18235210938136265,
+      "learning_rate": 2.2133436726966555e-06,
+      "loss": 0.5014,
+      "step": 918
+    },
+    {
+      "epoch": 0.8350749659245797,
+      "grad_norm": 0.18558447867998393,
+      "learning_rate": 2.189825162982648e-06,
+      "loss": 0.5052,
+      "step": 919
+    },
+    {
+      "epoch": 0.8359836437982735,
+      "grad_norm": 0.18104683947912312,
+      "learning_rate": 2.166422430162693e-06,
+      "loss": 0.4995,
+      "step": 920
+    },
+    {
+      "epoch": 0.8368923216719673,
+      "grad_norm": 0.18353267749631846,
+      "learning_rate": 2.1431356857480076e-06,
+      "loss": 0.4795,
+      "step": 921
+    },
+    {
+      "epoch": 0.8378009995456611,
+      "grad_norm": 0.19613272174120108,
+      "learning_rate": 2.1199651402015284e-06,
+      "loss": 0.5199,
+      "step": 922
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 0.19747489528791062,
+      "learning_rate": 2.0969110029359922e-06,
+      "loss": 0.5032,
+      "step": 923
+    },
+    {
+      "epoch": 0.8396183552930486,
+      "grad_norm": 0.1787049444368562,
+      "learning_rate": 2.073973482312051e-06,
+      "loss": 0.4967,
+      "step": 924
+    },
+    {
+      "epoch": 0.8405270331667424,
+      "grad_norm": 0.17606081747542848,
+      "learning_rate": 2.0511527856363916e-06,
+      "loss": 0.5041,
+      "step": 925
+    },
+    {
+      "epoch": 0.8414357110404361,
+      "grad_norm": 0.17829298424594497,
+      "learning_rate": 2.028449119159862e-06,
+      "loss": 0.5078,
+      "step": 926
+    },
+    {
+      "epoch": 0.84234438891413,
+      "grad_norm": 0.1948648000783994,
+      "learning_rate": 2.0058626880756014e-06,
+      "loss": 0.5072,
+      "step": 927
+    },
+    {
+      "epoch": 0.8432530667878237,
+      "grad_norm": 0.17927122485664945,
+      "learning_rate": 1.9833936965171817e-06,
+      "loss": 0.494,
+      "step": 928
+    },
+    {
+      "epoch": 0.8441617446615175,
+      "grad_norm": 0.1772461681235593,
+      "learning_rate": 1.961042347556787e-06,
+      "loss": 0.4923,
+      "step": 929
+    },
+    {
+      "epoch": 0.8450704225352113,
+      "grad_norm": 0.18305292727800654,
+      "learning_rate": 1.9388088432033446e-06,
+      "loss": 0.5261,
+      "step": 930
+    },
+    {
+      "epoch": 0.845979100408905,
+      "grad_norm": 0.1803832163167104,
+      "learning_rate": 1.916693384400722e-06,
+      "loss": 0.5219,
+      "step": 931
+    },
+    {
+      "epoch": 0.8468877782825989,
+      "grad_norm": 0.17898106251296597,
+      "learning_rate": 1.8946961710259048e-06,
+      "loss": 0.5111,
+      "step": 932
+    },
+    {
+      "epoch": 0.8477964561562926,
+      "grad_norm": 0.16947184822172653,
+      "learning_rate": 1.872817401887197e-06,
+      "loss": 0.5045,
+      "step": 933
+    },
+    {
+      "epoch": 0.8487051340299864,
+      "grad_norm": 0.17922642087683416,
+      "learning_rate": 1.8510572747224024e-06,
+      "loss": 0.4853,
+      "step": 934
+    },
+    {
+      "epoch": 0.8496138119036801,
+      "grad_norm": 0.17977834624667638,
+      "learning_rate": 1.8294159861970605e-06,
+      "loss": 0.5038,
+      "step": 935
+    },
+    {
+      "epoch": 0.8505224897773739,
+      "grad_norm": 0.19608404483438538,
+      "learning_rate": 1.8078937319026655e-06,
+      "loss": 0.5042,
+      "step": 936
+    },
+    {
+      "epoch": 0.8514311676510677,
+      "grad_norm": 0.1685912256631259,
+      "learning_rate": 1.7864907063548813e-06,
+      "loss": 0.5096,
+      "step": 937
+    },
+    {
+      "epoch": 0.8523398455247615,
+      "grad_norm": 0.17660218292868002,
+      "learning_rate": 1.7652071029918082e-06,
+      "loss": 0.4781,
+      "step": 938
+    },
+    {
+      "epoch": 0.8532485233984552,
+      "grad_norm": 0.18911622472242354,
+      "learning_rate": 1.7440431141722152e-06,
+      "loss": 0.5083,
+      "step": 939
+    },
+    {
+      "epoch": 0.854157201272149,
+      "grad_norm": 0.17455351806613617,
+      "learning_rate": 1.7229989311738142e-06,
+      "loss": 0.496,
+      "step": 940
+    },
+    {
+      "epoch": 0.8550658791458428,
+      "grad_norm": 0.1995844963464293,
+      "learning_rate": 1.7020747441915184e-06,
+      "loss": 0.502,
+      "step": 941
+    },
+    {
+      "epoch": 0.8559745570195366,
+      "grad_norm": 0.1668170199040863,
+      "learning_rate": 1.6812707423357387e-06,
+      "loss": 0.5095,
+      "step": 942
+    },
+    {
+      "epoch": 0.8568832348932304,
+      "grad_norm": 0.1948155057914624,
+      "learning_rate": 1.660587113630656e-06,
+      "loss": 0.4952,
+      "step": 943
+    },
+    {
+      "epoch": 0.8577919127669241,
+      "grad_norm": 0.18800602958995521,
+      "learning_rate": 1.6400240450125486e-06,
+      "loss": 0.4984,
+      "step": 944
+    },
+    {
+      "epoch": 0.8587005906406179,
+      "grad_norm": 0.17254706781068335,
+      "learning_rate": 1.619581722328073e-06,
+      "loss": 0.5151,
+      "step": 945
+    },
+    {
+      "epoch": 0.8596092685143116,
+      "grad_norm": 0.19375693939108612,
+      "learning_rate": 1.5992603303325997e-06,
+      "loss": 0.4976,
+      "step": 946
+    },
+    {
+      "epoch": 0.8605179463880055,
+      "grad_norm": 0.17558775095540655,
+      "learning_rate": 1.579060052688548e-06,
+      "loss": 0.5262,
+      "step": 947
+    },
+    {
+      "epoch": 0.8614266242616992,
+      "grad_norm": 0.17688437493103373,
+      "learning_rate": 1.5589810719637148e-06,
+      "loss": 0.507,
+      "step": 948
+    },
+    {
+      "epoch": 0.862335302135393,
+      "grad_norm": 0.18642075452138682,
+      "learning_rate": 1.5390235696296268e-06,
+      "loss": 0.4876,
+      "step": 949
+    },
+    {
+      "epoch": 0.8632439800090868,
+      "grad_norm": 0.17740686076054013,
+      "learning_rate": 1.519187726059903e-06,
+      "loss": 0.5048,
+      "step": 950
+    },
+    {
+      "epoch": 0.8641526578827805,
+      "grad_norm": 0.18919878445470903,
+      "learning_rate": 1.4994737205286375e-06,
+      "loss": 0.5152,
+      "step": 951
+    },
+    {
+      "epoch": 0.8650613357564744,
+      "grad_norm": 0.1907057801823801,
+      "learning_rate": 1.47988173120875e-06,
+      "loss": 0.5234,
+      "step": 952
+    },
+    {
+      "epoch": 0.8659700136301681,
+      "grad_norm": 0.1883206554410589,
+      "learning_rate": 1.4604119351703953e-06,
+      "loss": 0.5096,
+      "step": 953
+    },
+    {
+      "epoch": 0.8668786915038619,
+      "grad_norm": 0.17232570863058297,
+      "learning_rate": 1.441064508379366e-06,
+      "loss": 0.4783,
+      "step": 954
+    },
+    {
+      "epoch": 0.8677873693775556,
+      "grad_norm": 0.1762050568497747,
+      "learning_rate": 1.421839625695492e-06,
+      "loss": 0.4946,
+      "step": 955
+    },
+    {
+      "epoch": 0.8686960472512494,
+      "grad_norm": 0.1828489796685002,
+      "learning_rate": 1.402737460871057e-06,
+      "loss": 0.4925,
+      "step": 956
+    },
+    {
+      "epoch": 0.8696047251249432,
+      "grad_norm": 0.1874266302232557,
+      "learning_rate": 1.3837581865492428e-06,
+      "loss": 0.5129,
+      "step": 957
+    },
+    {
+      "epoch": 0.870513402998637,
+      "grad_norm": 0.17492796891040732,
+      "learning_rate": 1.3649019742625624e-06,
+      "loss": 0.4948,
+      "step": 958
+    },
+    {
+      "epoch": 0.8714220808723308,
+      "grad_norm": 0.17856469420360077,
+      "learning_rate": 1.346168994431302e-06,
+      "loss": 0.5047,
+      "step": 959
+    },
+    {
+      "epoch": 0.8723307587460245,
+      "grad_norm": 0.17735830612649864,
+      "learning_rate": 1.3275594163619875e-06,
+      "loss": 0.5119,
+      "step": 960
+    },
+    {
+      "epoch": 0.8732394366197183,
+      "grad_norm": 0.2000254861936507,
+      "learning_rate": 1.3090734082458562e-06,
+      "loss": 0.5315,
+      "step": 961
+    },
+    {
+      "epoch": 0.8741481144934121,
+      "grad_norm": 0.18755218675784746,
+      "learning_rate": 1.2907111371573356e-06,
+      "loss": 0.5117,
+      "step": 962
+    },
+    {
+      "epoch": 0.8750567923671059,
+      "grad_norm": 0.1924034058294698,
+      "learning_rate": 1.2724727690525278e-06,
+      "loss": 0.4963,
+      "step": 963
+    },
+    {
+      "epoch": 0.8759654702407996,
+      "grad_norm": 0.17199231210235086,
+      "learning_rate": 1.2543584687677163e-06,
+      "loss": 0.507,
+      "step": 964
+    },
+    {
+      "epoch": 0.8768741481144934,
+      "grad_norm": 0.195405734879678,
+      "learning_rate": 1.2363684000178788e-06,
+      "loss": 0.5061,
+      "step": 965
+    },
+    {
+      "epoch": 0.8777828259881872,
+      "grad_norm": 0.18129652277081798,
+      "learning_rate": 1.2185027253951935e-06,
+      "loss": 0.4902,
+      "step": 966
+    },
+    {
+      "epoch": 0.878691503861881,
+      "grad_norm": 0.17181670418876158,
+      "learning_rate": 1.200761606367583e-06,
+      "loss": 0.4705,
+      "step": 967
+    },
+    {
+      "epoch": 0.8796001817355747,
+      "grad_norm": 0.1782222124923922,
+      "learning_rate": 1.1831452032772499e-06,
+      "loss": 0.5054,
+      "step": 968
+    },
+    {
+      "epoch": 0.8805088596092685,
+      "grad_norm": 0.18034452027252415,
+      "learning_rate": 1.1656536753392288e-06,
+      "loss": 0.5068,
+      "step": 969
+    },
+    {
+      "epoch": 0.8814175374829623,
+      "grad_norm": 0.16897576149958596,
+      "learning_rate": 1.148287180639952e-06,
+      "loss": 0.5087,
+      "step": 970
+    },
+    {
+      "epoch": 0.8823262153566561,
+      "grad_norm": 0.17814376760510925,
+      "learning_rate": 1.1310458761358057e-06,
+      "loss": 0.5024,
+      "step": 971
+    },
+    {
+      "epoch": 0.8832348932303499,
+      "grad_norm": 0.1684038013793718,
+      "learning_rate": 1.113929917651731e-06,
+      "loss": 0.4988,
+      "step": 972
+    },
+    {
+      "epoch": 0.8841435711040436,
+      "grad_norm": 0.18680522710896053,
+      "learning_rate": 1.0969394598798005e-06,
+      "loss": 0.5146,
+      "step": 973
+    },
+    {
+      "epoch": 0.8850522489777374,
+      "grad_norm": 0.18089586831359786,
+      "learning_rate": 1.0800746563778258e-06,
+      "loss": 0.5051,
+      "step": 974
+    },
+    {
+      "epoch": 0.8859609268514311,
+      "grad_norm": 0.19317494093940246,
+      "learning_rate": 1.0633356595679683e-06,
+      "loss": 0.5025,
+      "step": 975
+    },
+    {
+      "epoch": 0.886869604725125,
+      "grad_norm": 0.27841178390301896,
+      "learning_rate": 1.0467226207353675e-06,
+      "loss": 0.4961,
+      "step": 976
+    },
+    {
+      "epoch": 0.8877782825988187,
+      "grad_norm": 0.1661933459945262,
+      "learning_rate": 1.03023569002677e-06,
+      "loss": 0.4928,
+      "step": 977
+    },
+    {
+      "epoch": 0.8886869604725125,
+      "grad_norm": 0.17343777056749088,
+      "learning_rate": 1.0138750164491678e-06,
+      "loss": 0.5091,
+      "step": 978
+    },
+    {
+      "epoch": 0.8895956383462063,
+      "grad_norm": 0.1889775529950286,
+      "learning_rate": 9.976407478684518e-07,
+      "loss": 0.4833,
+      "step": 979
+    },
+    {
+      "epoch": 0.8905043162199,
+      "grad_norm": 0.17817868099181625,
+      "learning_rate": 9.815330310080889e-07,
+      "loss": 0.4911,
+      "step": 980
+    },
+    {
+      "epoch": 0.8914129940935939,
+      "grad_norm": 0.16191917789368504,
+      "learning_rate": 9.655520114477772e-07,
+      "loss": 0.512,
+      "step": 981
+    },
+    {
+      "epoch": 0.8923216719672876,
+      "grad_norm": 0.18240957967746754,
+      "learning_rate": 9.496978336221396e-07,
+      "loss": 0.5045,
+      "step": 982
+    },
+    {
+      "epoch": 0.8932303498409814,
+      "grad_norm": 0.17017266671524892,
+      "learning_rate": 9.339706408194288e-07,
+      "loss": 0.4924,
+      "step": 983
+    },
+    {
+      "epoch": 0.8941390277146751,
+      "grad_norm": 0.17413211768706843,
+      "learning_rate": 9.183705751802074e-07,
+      "loss": 0.5049,
+      "step": 984
+    },
+    {
+      "epoch": 0.8950477055883689,
+      "grad_norm": 0.18605262758821256,
+      "learning_rate": 9.028977776960873e-07,
+      "loss": 0.5159,
+      "step": 985
+    },
+    {
+      "epoch": 0.8959563834620627,
+      "grad_norm": 0.19150551818628364,
+      "learning_rate": 8.875523882084352e-07,
+      "loss": 0.4868,
+      "step": 986
+    },
+    {
+      "epoch": 0.8968650613357565,
+      "grad_norm": 0.19088032763098828,
+      "learning_rate": 8.723345454071308e-07,
+      "loss": 0.5163,
+      "step": 987
+    },
+    {
+      "epoch": 0.8977737392094502,
+      "grad_norm": 0.16428972487904736,
+      "learning_rate": 8.57244386829294e-07,
+      "loss": 0.4949,
+      "step": 988
+    },
+    {
+      "epoch": 0.898682417083144,
+      "grad_norm": 0.16147474098681905,
+      "learning_rate": 8.422820488580462e-07,
+      "loss": 0.5111,
+      "step": 989
+    },
+    {
+      "epoch": 0.8995910949568378,
+      "grad_norm": 0.17877177293244242,
+      "learning_rate": 8.274476667212933e-07,
+      "loss": 0.4807,
+      "step": 990
+    },
+    {
+      "epoch": 0.9004997728305316,
+      "grad_norm": 0.17023754590293025,
+      "learning_rate": 8.127413744904805e-07,
+      "loss": 0.5124,
+      "step": 991
+    },
+    {
+      "epoch": 0.9014084507042254,
+      "grad_norm": 0.17706509947479346,
+      "learning_rate": 7.981633050793929e-07,
+      "loss": 0.4979,
+      "step": 992
+    },
+    {
+      "epoch": 0.9023171285779191,
+      "grad_norm": 0.18940926624622828,
+      "learning_rate": 7.837135902429509e-07,
+      "loss": 0.5098,
+      "step": 993
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 0.1689888104493502,
+      "learning_rate": 7.693923605760294e-07,
+      "loss": 0.511,
+      "step": 994
+    },
+    {
+      "epoch": 0.9041344843253066,
+      "grad_norm": 0.1736769118973061,
+      "learning_rate": 7.551997455122573e-07,
+      "loss": 0.4832,
+      "step": 995
+    },
+    {
+      "epoch": 0.9050431621990005,
+      "grad_norm": 0.16648584513031078,
+      "learning_rate": 7.411358733228679e-07,
+      "loss": 0.5054,
+      "step": 996
+    },
+    {
+      "epoch": 0.9059518400726942,
+      "grad_norm": 0.181474553011442,
+      "learning_rate": 7.272008711155292e-07,
+      "loss": 0.5086,
+      "step": 997
+    },
+    {
+      "epoch": 0.906860517946388,
+      "grad_norm": 0.17983163521623255,
+      "learning_rate": 7.133948648331956e-07,
+      "loss": 0.5099,
+      "step": 998
+    },
+    {
+      "epoch": 0.9077691958200818,
+      "grad_norm": 0.16423253255618706,
+      "learning_rate": 6.997179792529724e-07,
+      "loss": 0.4908,
+      "step": 999
+    },
+    {
+      "epoch": 0.9086778736937755,
+      "grad_norm": 0.17029107112544759,
+      "learning_rate": 6.861703379849815e-07,
+      "loss": 0.4954,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1100,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 976475935997952.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}