amburger66's picture
LoRA fine-tune on RobotSmith real world task00 ball in ring
11f786b verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.08,
"eval_steps": 50,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 8e-05,
"grad_norm": 28.904014587402344,
"learning_rate": 0.0,
"loss": 1.6623,
"step": 1
},
{
"epoch": 0.00016,
"grad_norm": 6.64036750793457,
"learning_rate": 2.0000000000000002e-07,
"loss": 0.7016,
"step": 2
},
{
"epoch": 0.00024,
"grad_norm": 7.72219705581665,
"learning_rate": 4.0000000000000003e-07,
"loss": 1.1014,
"step": 3
},
{
"epoch": 0.00032,
"grad_norm": 17.788040161132812,
"learning_rate": 6.000000000000001e-07,
"loss": 1.7937,
"step": 4
},
{
"epoch": 0.0004,
"grad_norm": 9.829057693481445,
"learning_rate": 8.000000000000001e-07,
"loss": 1.0861,
"step": 5
},
{
"epoch": 0.00048,
"grad_norm": 8.468535423278809,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.7452,
"step": 6
},
{
"epoch": 0.00056,
"grad_norm": 12.506383895874023,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.7238,
"step": 7
},
{
"epoch": 0.00064,
"grad_norm": 4.8890886306762695,
"learning_rate": 1.4000000000000001e-06,
"loss": 0.3983,
"step": 8
},
{
"epoch": 0.00072,
"grad_norm": 10.960965156555176,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.0239,
"step": 9
},
{
"epoch": 0.0008,
"grad_norm": 10.09985637664795,
"learning_rate": 1.8000000000000001e-06,
"loss": 1.0149,
"step": 10
},
{
"epoch": 0.00088,
"grad_norm": 17.378681182861328,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.0029,
"step": 11
},
{
"epoch": 0.00096,
"grad_norm": 7.184055328369141,
"learning_rate": 2.2e-06,
"loss": 0.8978,
"step": 12
},
{
"epoch": 0.00104,
"grad_norm": 11.876077651977539,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.9701,
"step": 13
},
{
"epoch": 0.00112,
"grad_norm": 15.245849609375,
"learning_rate": 2.6e-06,
"loss": 1.1197,
"step": 14
},
{
"epoch": 0.0012,
"grad_norm": 10.875532150268555,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.8404,
"step": 15
},
{
"epoch": 0.00128,
"grad_norm": 9.043937683105469,
"learning_rate": 3e-06,
"loss": 1.353,
"step": 16
},
{
"epoch": 0.00136,
"grad_norm": 9.204301834106445,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.7596,
"step": 17
},
{
"epoch": 0.00144,
"grad_norm": 7.705916881561279,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.8459,
"step": 18
},
{
"epoch": 0.00152,
"grad_norm": 11.356510162353516,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.4476,
"step": 19
},
{
"epoch": 0.0016,
"grad_norm": 7.212140083312988,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.7956,
"step": 20
},
{
"epoch": 0.00168,
"grad_norm": 22.89238166809082,
"learning_rate": 4.000000000000001e-06,
"loss": 1.4061,
"step": 21
},
{
"epoch": 0.00176,
"grad_norm": 7.023189544677734,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.6018,
"step": 22
},
{
"epoch": 0.00184,
"grad_norm": 11.999281883239746,
"learning_rate": 4.4e-06,
"loss": 1.0959,
"step": 23
},
{
"epoch": 0.00192,
"grad_norm": 8.107518196105957,
"learning_rate": 4.600000000000001e-06,
"loss": 0.415,
"step": 24
},
{
"epoch": 0.002,
"grad_norm": 10.710831642150879,
"learning_rate": 4.800000000000001e-06,
"loss": 1.099,
"step": 25
},
{
"epoch": 0.00208,
"grad_norm": 9.119357109069824,
"learning_rate": 5e-06,
"loss": 0.7439,
"step": 26
},
{
"epoch": 0.00216,
"grad_norm": 10.674591064453125,
"learning_rate": 5.2e-06,
"loss": 0.693,
"step": 27
},
{
"epoch": 0.00224,
"grad_norm": 5.260153293609619,
"learning_rate": 5.400000000000001e-06,
"loss": 0.3789,
"step": 28
},
{
"epoch": 0.00232,
"grad_norm": 15.196737289428711,
"learning_rate": 5.600000000000001e-06,
"loss": 0.9968,
"step": 29
},
{
"epoch": 0.0024,
"grad_norm": 13.141084671020508,
"learning_rate": 5.8e-06,
"loss": 1.1159,
"step": 30
},
{
"epoch": 0.00248,
"grad_norm": 9.105382919311523,
"learning_rate": 6e-06,
"loss": 0.5438,
"step": 31
},
{
"epoch": 0.00256,
"grad_norm": 10.478144645690918,
"learning_rate": 6.200000000000001e-06,
"loss": 0.9363,
"step": 32
},
{
"epoch": 0.00264,
"grad_norm": 9.590822219848633,
"learning_rate": 6.4000000000000006e-06,
"loss": 0.7012,
"step": 33
},
{
"epoch": 0.00272,
"grad_norm": 6.54506254196167,
"learning_rate": 6.600000000000001e-06,
"loss": 0.9276,
"step": 34
},
{
"epoch": 0.0028,
"grad_norm": 11.808697700500488,
"learning_rate": 6.800000000000001e-06,
"loss": 0.9077,
"step": 35
},
{
"epoch": 0.00288,
"grad_norm": 11.456603050231934,
"learning_rate": 7e-06,
"loss": 0.8894,
"step": 36
},
{
"epoch": 0.00296,
"grad_norm": 11.408453941345215,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.7536,
"step": 37
},
{
"epoch": 0.00304,
"grad_norm": 4.814957618713379,
"learning_rate": 7.4e-06,
"loss": 0.3702,
"step": 38
},
{
"epoch": 0.00312,
"grad_norm": 8.178116798400879,
"learning_rate": 7.600000000000001e-06,
"loss": 0.5176,
"step": 39
},
{
"epoch": 0.0032,
"grad_norm": 11.304093360900879,
"learning_rate": 7.800000000000002e-06,
"loss": 0.8417,
"step": 40
},
{
"epoch": 0.00328,
"grad_norm": 19.103084564208984,
"learning_rate": 8.000000000000001e-06,
"loss": 0.9916,
"step": 41
},
{
"epoch": 0.00336,
"grad_norm": 13.37364387512207,
"learning_rate": 8.2e-06,
"loss": 0.8813,
"step": 42
},
{
"epoch": 0.00344,
"grad_norm": 4.601191997528076,
"learning_rate": 8.400000000000001e-06,
"loss": 0.4729,
"step": 43
},
{
"epoch": 0.00352,
"grad_norm": 6.08097505569458,
"learning_rate": 8.6e-06,
"loss": 0.2231,
"step": 44
},
{
"epoch": 0.0036,
"grad_norm": 6.659447193145752,
"learning_rate": 8.8e-06,
"loss": 0.4153,
"step": 45
},
{
"epoch": 0.00368,
"grad_norm": 11.243528366088867,
"learning_rate": 9e-06,
"loss": 0.3482,
"step": 46
},
{
"epoch": 0.00376,
"grad_norm": 9.115594863891602,
"learning_rate": 9.200000000000002e-06,
"loss": 0.7936,
"step": 47
},
{
"epoch": 0.00384,
"grad_norm": 9.980374336242676,
"learning_rate": 9.4e-06,
"loss": 0.6314,
"step": 48
},
{
"epoch": 0.00392,
"grad_norm": 7.791968822479248,
"learning_rate": 9.600000000000001e-06,
"loss": 0.6718,
"step": 49
},
{
"epoch": 0.004,
"grad_norm": 3.5061657428741455,
"learning_rate": 9.800000000000001e-06,
"loss": 0.3744,
"step": 50
},
{
"epoch": 0.00408,
"grad_norm": 5.280981540679932,
"learning_rate": 1e-05,
"loss": 0.3512,
"step": 51
},
{
"epoch": 0.00416,
"grad_norm": 5.641449451446533,
"learning_rate": 1.02e-05,
"loss": 0.4139,
"step": 52
},
{
"epoch": 0.00424,
"grad_norm": 4.7260637283325195,
"learning_rate": 1.04e-05,
"loss": 0.3731,
"step": 53
},
{
"epoch": 0.00432,
"grad_norm": 6.269501209259033,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.5638,
"step": 54
},
{
"epoch": 0.0044,
"grad_norm": 2.2564549446105957,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.2487,
"step": 55
},
{
"epoch": 0.00448,
"grad_norm": 5.405374050140381,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.376,
"step": 56
},
{
"epoch": 0.00456,
"grad_norm": 8.047243118286133,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.6199,
"step": 57
},
{
"epoch": 0.00464,
"grad_norm": 9.34673023223877,
"learning_rate": 1.14e-05,
"loss": 0.5024,
"step": 58
},
{
"epoch": 0.00472,
"grad_norm": 6.233428955078125,
"learning_rate": 1.16e-05,
"loss": 0.451,
"step": 59
},
{
"epoch": 0.0048,
"grad_norm": 2.954914093017578,
"learning_rate": 1.18e-05,
"loss": 0.2806,
"step": 60
},
{
"epoch": 0.00488,
"grad_norm": 5.369543075561523,
"learning_rate": 1.2e-05,
"loss": 0.3224,
"step": 61
},
{
"epoch": 0.00496,
"grad_norm": 2.8051111698150635,
"learning_rate": 1.22e-05,
"loss": 0.3623,
"step": 62
},
{
"epoch": 0.00504,
"grad_norm": 3.5070888996124268,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.286,
"step": 63
},
{
"epoch": 0.00512,
"grad_norm": 2.8483386039733887,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.1474,
"step": 64
},
{
"epoch": 0.0052,
"grad_norm": 5.642381191253662,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.4356,
"step": 65
},
{
"epoch": 0.00528,
"grad_norm": 9.534383773803711,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.4917,
"step": 66
},
{
"epoch": 0.00536,
"grad_norm": 4.600446701049805,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.299,
"step": 67
},
{
"epoch": 0.00544,
"grad_norm": 3.365933418273926,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.0782,
"step": 68
},
{
"epoch": 0.00552,
"grad_norm": 2.059140920639038,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.1978,
"step": 69
},
{
"epoch": 0.0056,
"grad_norm": 6.936392307281494,
"learning_rate": 1.38e-05,
"loss": 0.4749,
"step": 70
},
{
"epoch": 0.00568,
"grad_norm": 6.204529285430908,
"learning_rate": 1.4e-05,
"loss": 0.4281,
"step": 71
},
{
"epoch": 0.00576,
"grad_norm": 9.302611351013184,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.6317,
"step": 72
},
{
"epoch": 0.00584,
"grad_norm": 3.6891136169433594,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.2884,
"step": 73
},
{
"epoch": 0.00592,
"grad_norm": 3.9807183742523193,
"learning_rate": 1.46e-05,
"loss": 0.3235,
"step": 74
},
{
"epoch": 0.006,
"grad_norm": 1.887704610824585,
"learning_rate": 1.48e-05,
"loss": 0.2188,
"step": 75
},
{
"epoch": 0.00608,
"grad_norm": 2.5952062606811523,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.2232,
"step": 76
},
{
"epoch": 0.00616,
"grad_norm": 5.21189022064209,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.5987,
"step": 77
},
{
"epoch": 0.00624,
"grad_norm": 2.4845821857452393,
"learning_rate": 1.54e-05,
"loss": 0.259,
"step": 78
},
{
"epoch": 0.00632,
"grad_norm": 2.9473681449890137,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.3172,
"step": 79
},
{
"epoch": 0.0064,
"grad_norm": 3.7112972736358643,
"learning_rate": 1.58e-05,
"loss": 0.281,
"step": 80
},
{
"epoch": 0.00648,
"grad_norm": 6.32492733001709,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.445,
"step": 81
},
{
"epoch": 0.00656,
"grad_norm": 3.0493485927581787,
"learning_rate": 1.62e-05,
"loss": 0.2145,
"step": 82
},
{
"epoch": 0.00664,
"grad_norm": 18.866405487060547,
"learning_rate": 1.64e-05,
"loss": 0.7711,
"step": 83
},
{
"epoch": 0.00672,
"grad_norm": 0.8289015889167786,
"learning_rate": 1.66e-05,
"loss": 0.0183,
"step": 84
},
{
"epoch": 0.0068,
"grad_norm": 4.44669771194458,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.2556,
"step": 85
},
{
"epoch": 0.00688,
"grad_norm": 1.3111125230789185,
"learning_rate": 1.7e-05,
"loss": 0.1472,
"step": 86
},
{
"epoch": 0.00696,
"grad_norm": 7.138492584228516,
"learning_rate": 1.72e-05,
"loss": 0.4512,
"step": 87
},
{
"epoch": 0.00704,
"grad_norm": 2.2903518676757812,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.2546,
"step": 88
},
{
"epoch": 0.00712,
"grad_norm": 1.3174657821655273,
"learning_rate": 1.76e-05,
"loss": 0.1638,
"step": 89
},
{
"epoch": 0.0072,
"grad_norm": 1.0058629512786865,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.0339,
"step": 90
},
{
"epoch": 0.00728,
"grad_norm": 8.9186429977417,
"learning_rate": 1.8e-05,
"loss": 0.2801,
"step": 91
},
{
"epoch": 0.00736,
"grad_norm": 5.189001560211182,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.1255,
"step": 92
},
{
"epoch": 0.00744,
"grad_norm": 6.905643939971924,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.38,
"step": 93
},
{
"epoch": 0.00752,
"grad_norm": 2.7954299449920654,
"learning_rate": 1.86e-05,
"loss": 0.1494,
"step": 94
},
{
"epoch": 0.0076,
"grad_norm": 2.9016220569610596,
"learning_rate": 1.88e-05,
"loss": 0.2209,
"step": 95
},
{
"epoch": 0.00768,
"grad_norm": 3.5789694786071777,
"learning_rate": 1.9e-05,
"loss": 0.1116,
"step": 96
},
{
"epoch": 0.00776,
"grad_norm": 2.2798409461975098,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.0536,
"step": 97
},
{
"epoch": 0.00784,
"grad_norm": 1.73595130443573,
"learning_rate": 1.94e-05,
"loss": 0.1721,
"step": 98
},
{
"epoch": 0.00792,
"grad_norm": 7.032018661499023,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.2286,
"step": 99
},
{
"epoch": 0.008,
"grad_norm": 2.365339756011963,
"learning_rate": 1.98e-05,
"loss": 0.1825,
"step": 100
},
{
"epoch": 0.00808,
"grad_norm": 1.2685211896896362,
"learning_rate": 2e-05,
"loss": 0.0421,
"step": 101
},
{
"epoch": 0.00816,
"grad_norm": 1.1217608451843262,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.1293,
"step": 102
},
{
"epoch": 0.00824,
"grad_norm": 2.50956392288208,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.2329,
"step": 103
},
{
"epoch": 0.00832,
"grad_norm": 3.357452154159546,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.1515,
"step": 104
},
{
"epoch": 0.0084,
"grad_norm": 2.4834811687469482,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.2058,
"step": 105
},
{
"epoch": 0.00848,
"grad_norm": 7.074051380157471,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.3445,
"step": 106
},
{
"epoch": 0.00856,
"grad_norm": 13.237238883972168,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.3909,
"step": 107
},
{
"epoch": 0.00864,
"grad_norm": 1.1928606033325195,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.1277,
"step": 108
},
{
"epoch": 0.00872,
"grad_norm": 1.208445429801941,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.1132,
"step": 109
},
{
"epoch": 0.0088,
"grad_norm": 2.0769426822662354,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.1662,
"step": 110
},
{
"epoch": 0.00888,
"grad_norm": 2.3081891536712646,
"learning_rate": 1.999390827019096e-05,
"loss": 0.1812,
"step": 111
},
{
"epoch": 0.00896,
"grad_norm": 3.6210570335388184,
"learning_rate": 1.999262916410621e-05,
"loss": 0.1635,
"step": 112
},
{
"epoch": 0.00904,
"grad_norm": 12.55403995513916,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.3269,
"step": 113
},
{
"epoch": 0.00912,
"grad_norm": 1.4872065782546997,
"learning_rate": 1.998970569790715e-05,
"loss": 0.1024,
"step": 114
},
{
"epoch": 0.0092,
"grad_norm": 7.752160549163818,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.4973,
"step": 115
},
{
"epoch": 0.00928,
"grad_norm": 11.699462890625,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.5843,
"step": 116
},
{
"epoch": 0.00936,
"grad_norm": 1.1619901657104492,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.1332,
"step": 117
},
{
"epoch": 0.00944,
"grad_norm": 1.5630009174346924,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.1987,
"step": 118
},
{
"epoch": 0.00952,
"grad_norm": 1.3593034744262695,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.1159,
"step": 119
},
{
"epoch": 0.0096,
"grad_norm": 10.937616348266602,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.4075,
"step": 120
},
{
"epoch": 0.00968,
"grad_norm": 2.624112367630005,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.2251,
"step": 121
},
{
"epoch": 0.00976,
"grad_norm": 4.521087169647217,
"learning_rate": 1.997314477224458e-05,
"loss": 0.3206,
"step": 122
},
{
"epoch": 0.00984,
"grad_norm": 2.7348177433013916,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.2106,
"step": 123
},
{
"epoch": 0.00992,
"grad_norm": 2.018728494644165,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.1813,
"step": 124
},
{
"epoch": 0.01,
"grad_norm": 1.5363668203353882,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.1231,
"step": 125
},
{
"epoch": 0.01008,
"grad_norm": 1.4855091571807861,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.059,
"step": 126
},
{
"epoch": 0.01016,
"grad_norm": 2.038907527923584,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.3014,
"step": 127
},
{
"epoch": 0.01024,
"grad_norm": 2.4306297302246094,
"learning_rate": 1.99556196460308e-05,
"loss": 0.1949,
"step": 128
},
{
"epoch": 0.01032,
"grad_norm": 1.051932454109192,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.1146,
"step": 129
},
{
"epoch": 0.0104,
"grad_norm": 1.5948859453201294,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.1388,
"step": 130
},
{
"epoch": 0.01048,
"grad_norm": 0.7300507426261902,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.0943,
"step": 131
},
{
"epoch": 0.01056,
"grad_norm": 3.148203134536743,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.2338,
"step": 132
},
{
"epoch": 0.01064,
"grad_norm": 2.271157741546631,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.212,
"step": 133
},
{
"epoch": 0.01072,
"grad_norm": 0.632797360420227,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.1162,
"step": 134
},
{
"epoch": 0.0108,
"grad_norm": 13.224139213562012,
"learning_rate": 1.992965508106537e-05,
"loss": 0.6016,
"step": 135
},
{
"epoch": 0.01088,
"grad_norm": 1.4903218746185303,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.15,
"step": 136
},
{
"epoch": 0.01096,
"grad_norm": 1.6475129127502441,
"learning_rate": 1.9921147013144782e-05,
"loss": 0.1551,
"step": 137
},
{
"epoch": 0.01104,
"grad_norm": 7.009321212768555,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.1942,
"step": 138
},
{
"epoch": 0.01112,
"grad_norm": 1.0233148336410522,
"learning_rate": 1.991215540251542e-05,
"loss": 0.114,
"step": 139
},
{
"epoch": 0.0112,
"grad_norm": 2.4258909225463867,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.1871,
"step": 140
},
{
"epoch": 0.01128,
"grad_norm": 2.5366311073303223,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.1803,
"step": 141
},
{
"epoch": 0.01136,
"grad_norm": 11.73539924621582,
"learning_rate": 1.989776230907789e-05,
"loss": 0.6668,
"step": 142
},
{
"epoch": 0.01144,
"grad_norm": 0.8502317070960999,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.1094,
"step": 143
},
{
"epoch": 0.01152,
"grad_norm": 1.354844331741333,
"learning_rate": 1.988756381047006e-05,
"loss": 0.1218,
"step": 144
},
{
"epoch": 0.0116,
"grad_norm": 0.7307093143463135,
"learning_rate": 1.988228381446553e-05,
"loss": 0.1137,
"step": 145
},
{
"epoch": 0.01168,
"grad_norm": 2.2051281929016113,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.1297,
"step": 146
},
{
"epoch": 0.01176,
"grad_norm": 0.9884032607078552,
"learning_rate": 1.987136265072988e-05,
"loss": 0.1074,
"step": 147
},
{
"epoch": 0.01184,
"grad_norm": 1.0043885707855225,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.123,
"step": 148
},
{
"epoch": 0.01192,
"grad_norm": 2.3129684925079346,
"learning_rate": 1.985996037070505e-05,
"loss": 0.1474,
"step": 149
},
{
"epoch": 0.012,
"grad_norm": 11.889053344726562,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.2875,
"step": 150
},
{
"epoch": 0.01208,
"grad_norm": 1.2879987955093384,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.1326,
"step": 151
},
{
"epoch": 0.01216,
"grad_norm": 0.9905913472175598,
"learning_rate": 1.984195607969242e-05,
"loss": 0.104,
"step": 152
},
{
"epoch": 0.01224,
"grad_norm": 3.7368478775024414,
"learning_rate": 1.983571470813386e-05,
"loss": 0.2819,
"step": 153
},
{
"epoch": 0.01232,
"grad_norm": 6.304781436920166,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.2724,
"step": 154
},
{
"epoch": 0.0124,
"grad_norm": 1.3073478937149048,
"learning_rate": 1.982287250728689e-05,
"loss": 0.1561,
"step": 155
},
{
"epoch": 0.01248,
"grad_norm": 12.779458045959473,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.5108,
"step": 156
},
{
"epoch": 0.01256,
"grad_norm": 1.2683457136154175,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.2002,
"step": 157
},
{
"epoch": 0.01264,
"grad_norm": 0.6664188504219055,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.0138,
"step": 158
},
{
"epoch": 0.01272,
"grad_norm": 0.4715230464935303,
"learning_rate": 1.979575249599344e-05,
"loss": 0.0092,
"step": 159
},
{
"epoch": 0.0128,
"grad_norm": 9.394930839538574,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.3839,
"step": 160
},
{
"epoch": 0.01288,
"grad_norm": 1.8819639682769775,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.1208,
"step": 161
},
{
"epoch": 0.01296,
"grad_norm": 5.742199420928955,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.2433,
"step": 162
},
{
"epoch": 0.01304,
"grad_norm": 2.501323938369751,
"learning_rate": 1.9766722783341682e-05,
"loss": 0.1361,
"step": 163
},
{
"epoch": 0.01312,
"grad_norm": 1.7430888414382935,
"learning_rate": 1.9759167619387474e-05,
"loss": 0.0504,
"step": 164
},
{
"epoch": 0.0132,
"grad_norm": 3.5373475551605225,
"learning_rate": 1.9751493543055634e-05,
"loss": 0.2441,
"step": 165
},
{
"epoch": 0.01328,
"grad_norm": 3.328591823577881,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.1988,
"step": 166
},
{
"epoch": 0.01336,
"grad_norm": 2.908790111541748,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.1799,
"step": 167
},
{
"epoch": 0.01344,
"grad_norm": 1.544342041015625,
"learning_rate": 1.972775878209397e-05,
"loss": 0.1184,
"step": 168
},
{
"epoch": 0.01352,
"grad_norm": 4.8449625968933105,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.3952,
"step": 169
},
{
"epoch": 0.0136,
"grad_norm": 0.5794314742088318,
"learning_rate": 1.971134279909636e-05,
"loss": 0.1032,
"step": 170
},
{
"epoch": 0.01368,
"grad_norm": 1.5584802627563477,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.1277,
"step": 171
},
{
"epoch": 0.01376,
"grad_norm": 1.1445811986923218,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.1301,
"step": 172
},
{
"epoch": 0.01384,
"grad_norm": 1.5739485025405884,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.106,
"step": 173
},
{
"epoch": 0.01392,
"grad_norm": 6.968087196350098,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.4243,
"step": 174
},
{
"epoch": 0.014,
"grad_norm": 0.7585766315460205,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.0801,
"step": 175
},
{
"epoch": 0.01408,
"grad_norm": 1.6557797193527222,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.0998,
"step": 176
},
{
"epoch": 0.01416,
"grad_norm": 1.190155029296875,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.035,
"step": 177
},
{
"epoch": 0.01424,
"grad_norm": 1.2722744941711426,
"learning_rate": 1.96409540423411e-05,
"loss": 0.1176,
"step": 178
},
{
"epoch": 0.01432,
"grad_norm": 1.855873465538025,
"learning_rate": 1.9631625667976584e-05,
"loss": 0.1235,
"step": 179
},
{
"epoch": 0.0144,
"grad_norm": 1.4007898569107056,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.1077,
"step": 180
},
{
"epoch": 0.01448,
"grad_norm": 2.2819056510925293,
"learning_rate": 1.961261695938319e-05,
"loss": 0.116,
"step": 181
},
{
"epoch": 0.01456,
"grad_norm": 2.311680555343628,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.1169,
"step": 182
},
{
"epoch": 0.01464,
"grad_norm": 0.20164479315280914,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.0005,
"step": 183
},
{
"epoch": 0.01472,
"grad_norm": 0.8525561094284058,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.0977,
"step": 184
},
{
"epoch": 0.0148,
"grad_norm": 1.9335949420928955,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.0417,
"step": 185
},
{
"epoch": 0.01488,
"grad_norm": 1.4592931270599365,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.1131,
"step": 186
},
{
"epoch": 0.01496,
"grad_norm": 1.0293995141983032,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.0887,
"step": 187
},
{
"epoch": 0.01504,
"grad_norm": 1.7292391061782837,
"learning_rate": 1.954240328516277e-05,
"loss": 0.1441,
"step": 188
},
{
"epoch": 0.01512,
"grad_norm": 2.778445243835449,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.1342,
"step": 189
},
{
"epoch": 0.0152,
"grad_norm": 0.977188766002655,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.094,
"step": 190
},
{
"epoch": 0.01528,
"grad_norm": 0.4342818558216095,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.0688,
"step": 191
},
{
"epoch": 0.01536,
"grad_norm": 1.7098332643508911,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.2007,
"step": 192
},
{
"epoch": 0.01544,
"grad_norm": 0.7625004649162292,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.0952,
"step": 193
},
{
"epoch": 0.01552,
"grad_norm": 0.7774609327316284,
"learning_rate": 1.947768410009586e-05,
"loss": 0.0843,
"step": 194
},
{
"epoch": 0.0156,
"grad_norm": 0.6080492734909058,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.0767,
"step": 195
},
{
"epoch": 0.01568,
"grad_norm": 0.0836167186498642,
"learning_rate": 1.945518575599317e-05,
"loss": 0.0008,
"step": 196
},
{
"epoch": 0.01576,
"grad_norm": 0.7541863322257996,
"learning_rate": 1.944376370237481e-05,
"loss": 0.0885,
"step": 197
},
{
"epoch": 0.01584,
"grad_norm": 2.8570916652679443,
"learning_rate": 1.943222657947601e-05,
"loss": 0.1141,
"step": 198
},
{
"epoch": 0.01592,
"grad_norm": 0.7610045075416565,
"learning_rate": 1.942057452787297e-05,
"loss": 0.0896,
"step": 199
},
{
"epoch": 0.016,
"grad_norm": 0.5514517426490784,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.0821,
"step": 200
},
{
"epoch": 0.01608,
"grad_norm": 1.305419683456421,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.2319,
"step": 201
},
{
"epoch": 0.01616,
"grad_norm": 0.6161731481552124,
"learning_rate": 1.938493022759556e-05,
"loss": 0.0963,
"step": 202
},
{
"epoch": 0.01624,
"grad_norm": 1.3213708400726318,
"learning_rate": 1.937281989491892e-05,
"loss": 0.0127,
"step": 203
},
{
"epoch": 0.01632,
"grad_norm": 0.6355310082435608,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.0914,
"step": 204
},
{
"epoch": 0.0164,
"grad_norm": 0.6373775601387024,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.1066,
"step": 205
},
{
"epoch": 0.01648,
"grad_norm": 0.7500351667404175,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.0801,
"step": 206
},
{
"epoch": 0.01656,
"grad_norm": 0.909795880317688,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.0754,
"step": 207
},
{
"epoch": 0.01664,
"grad_norm": 0.5503889918327332,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.0727,
"step": 208
},
{
"epoch": 0.01672,
"grad_norm": 0.8646226525306702,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.0926,
"step": 209
},
{
"epoch": 0.0168,
"grad_norm": 19.88033676147461,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.4731,
"step": 210
},
{
"epoch": 0.01688,
"grad_norm": 0.7570164203643799,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.0827,
"step": 211
},
{
"epoch": 0.01696,
"grad_norm": 0.950798511505127,
"learning_rate": 1.925870584809995e-05,
"loss": 0.0937,
"step": 212
},
{
"epoch": 0.01704,
"grad_norm": 0.19956669211387634,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.003,
"step": 213
},
{
"epoch": 0.01712,
"grad_norm": 1.6361727714538574,
"learning_rate": 1.923210217112981e-05,
"loss": 0.0996,
"step": 214
},
{
"epoch": 0.0172,
"grad_norm": 0.9541513323783875,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.0906,
"step": 215
},
{
"epoch": 0.01728,
"grad_norm": 2.612729787826538,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.1444,
"step": 216
},
{
"epoch": 0.01736,
"grad_norm": 0.1337594836950302,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.0016,
"step": 217
},
{
"epoch": 0.01744,
"grad_norm": 0.5602476596832275,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.0696,
"step": 218
},
{
"epoch": 0.01752,
"grad_norm": 2.235830545425415,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.1139,
"step": 219
},
{
"epoch": 0.0176,
"grad_norm": 0.566162645816803,
"learning_rate": 1.914959667849825e-05,
"loss": 0.0716,
"step": 220
},
{
"epoch": 0.01768,
"grad_norm": 0.6070978045463562,
"learning_rate": 1.913545457642601e-05,
"loss": 0.0758,
"step": 221
},
{
"epoch": 0.01776,
"grad_norm": 0.4532201886177063,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.0723,
"step": 222
},
{
"epoch": 0.01784,
"grad_norm": 1.956010341644287,
"learning_rate": 1.910683660806177e-05,
"loss": 0.0844,
"step": 223
},
{
"epoch": 0.01792,
"grad_norm": 14.386972427368164,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.2973,
"step": 224
},
{
"epoch": 0.018,
"grad_norm": 1.0450690984725952,
"learning_rate": 1.907777478532909e-05,
"loss": 0.0897,
"step": 225
},
{
"epoch": 0.01808,
"grad_norm": 0.7512031197547913,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.0138,
"step": 226
},
{
"epoch": 0.01816,
"grad_norm": 0.6857627630233765,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.0775,
"step": 227
},
{
"epoch": 0.01824,
"grad_norm": 1.0806868076324463,
"learning_rate": 1.903335292863301e-05,
"loss": 0.1059,
"step": 228
},
{
"epoch": 0.01832,
"grad_norm": 2.642319440841675,
"learning_rate": 1.901832526405114e-05,
"loss": 0.243,
"step": 229
},
{
"epoch": 0.0184,
"grad_norm": 1.021954894065857,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.0869,
"step": 230
},
{
"epoch": 0.01848,
"grad_norm": 0.9021756649017334,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.0845,
"step": 231
},
{
"epoch": 0.01856,
"grad_norm": 2.749837875366211,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.1396,
"step": 232
},
{
"epoch": 0.01864,
"grad_norm": 1.3611373901367188,
"learning_rate": 1.895711760239413e-05,
"loss": 0.0903,
"step": 233
},
{
"epoch": 0.01872,
"grad_norm": 0.794328510761261,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.0724,
"step": 234
},
{
"epoch": 0.0188,
"grad_norm": 0.38149958848953247,
"learning_rate": 1.892585818452126e-05,
"loss": 0.0553,
"step": 235
},
{
"epoch": 0.01888,
"grad_norm": 1.1677418947219849,
"learning_rate": 1.891006524188368e-05,
"loss": 0.1131,
"step": 236
},
{
"epoch": 0.01896,
"grad_norm": 0.5948715806007385,
"learning_rate": 1.889416373291298e-05,
"loss": 0.0445,
"step": 237
},
{
"epoch": 0.01904,
"grad_norm": 1.6000583171844482,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.133,
"step": 238
},
{
"epoch": 0.01912,
"grad_norm": 0.7666167616844177,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.0724,
"step": 239
},
{
"epoch": 0.0192,
"grad_norm": 0.48546305298805237,
"learning_rate": 1.884580975215084e-05,
"loss": 0.0747,
"step": 240
},
{
"epoch": 0.01928,
"grad_norm": 1.9748789072036743,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.162,
"step": 241
},
{
"epoch": 0.01936,
"grad_norm": 0.3267890214920044,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.0054,
"step": 242
},
{
"epoch": 0.01944,
"grad_norm": 0.8022861480712891,
"learning_rate": 1.879648572866617e-05,
"loss": 0.0712,
"step": 243
},
{
"epoch": 0.01952,
"grad_norm": 0.7736788392066956,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.0686,
"step": 244
},
{
"epoch": 0.0196,
"grad_norm": 0.0644134134054184,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.0011,
"step": 245
},
{
"epoch": 0.01968,
"grad_norm": 0.6066959500312805,
"learning_rate": 1.874619707139396e-05,
"loss": 0.0758,
"step": 246
},
{
"epoch": 0.01976,
"grad_norm": 1.4355800151824951,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.109,
"step": 247
},
{
"epoch": 0.01984,
"grad_norm": 0.11132998764514923,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.0015,
"step": 248
},
{
"epoch": 0.01992,
"grad_norm": 4.063149452209473,
"learning_rate": 1.869494929505219e-05,
"loss": 0.1211,
"step": 249
},
{
"epoch": 0.02,
"grad_norm": 1.7751275300979614,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.121,
"step": 250
},
{
"epoch": 0.02008,
"grad_norm": 0.44498610496520996,
"learning_rate": 1.866025403784439e-05,
"loss": 0.0718,
"step": 251
},
{
"epoch": 0.02016,
"grad_norm": 0.654429018497467,
"learning_rate": 1.864274801953705e-05,
"loss": 0.0728,
"step": 252
},
{
"epoch": 0.02024,
"grad_norm": 0.8691314458847046,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.0746,
"step": 253
},
{
"epoch": 0.02032,
"grad_norm": 0.2548674941062927,
"learning_rate": 1.860742027003944e-05,
"loss": 0.005,
"step": 254
},
{
"epoch": 0.0204,
"grad_norm": 0.9602722525596619,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.0901,
"step": 255
},
{
"epoch": 0.02048,
"grad_norm": 1.002867341041565,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.0715,
"step": 256
},
{
"epoch": 0.02056,
"grad_norm": 0.5777568221092224,
"learning_rate": 1.855364260160507e-05,
"loss": 0.0626,
"step": 257
},
{
"epoch": 0.02064,
"grad_norm": 0.9667588472366333,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.0888,
"step": 258
},
{
"epoch": 0.02072,
"grad_norm": 0.927061140537262,
"learning_rate": 1.851726934143048e-05,
"loss": 0.1095,
"step": 259
},
{
"epoch": 0.0208,
"grad_norm": 6.399965286254883,
"learning_rate": 1.849892692986864e-05,
"loss": 0.205,
"step": 260
},
{
"epoch": 0.02088,
"grad_norm": 1.1082454919815063,
"learning_rate": 1.848048096156426e-05,
"loss": 0.128,
"step": 261
},
{
"epoch": 0.02096,
"grad_norm": 0.3593654930591583,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.0551,
"step": 262
},
{
"epoch": 0.02104,
"grad_norm": 1.4085252285003662,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.0903,
"step": 263
},
{
"epoch": 0.02112,
"grad_norm": 1.152054786682129,
"learning_rate": 1.842452397007148e-05,
"loss": 0.0781,
"step": 264
},
{
"epoch": 0.0212,
"grad_norm": 0.9289559125900269,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.1068,
"step": 265
},
{
"epoch": 0.02128,
"grad_norm": 0.5249579548835754,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.0732,
"step": 266
},
{
"epoch": 0.02136,
"grad_norm": 0.8387550711631775,
"learning_rate": 1.836764313458962e-05,
"loss": 0.1114,
"step": 267
},
{
"epoch": 0.02144,
"grad_norm": 1.0738202333450317,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.106,
"step": 268
},
{
"epoch": 0.02152,
"grad_norm": 0.7749744653701782,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.0735,
"step": 269
},
{
"epoch": 0.0216,
"grad_norm": 0.7427696585655212,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.0541,
"step": 270
},
{
"epoch": 0.02168,
"grad_norm": 1.256883144378662,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.1182,
"step": 271
},
{
"epoch": 0.02176,
"grad_norm": 1.2820348739624023,
"learning_rate": 1.827080574274562e-05,
"loss": 0.0752,
"step": 272
},
{
"epoch": 0.02184,
"grad_norm": 13.88135051727295,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.4837,
"step": 273
},
{
"epoch": 0.02192,
"grad_norm": 1.0842207670211792,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.0722,
"step": 274
},
{
"epoch": 0.022,
"grad_norm": 1.6404986381530762,
"learning_rate": 1.821149209133704e-05,
"loss": 0.085,
"step": 275
},
{
"epoch": 0.02208,
"grad_norm": 2.777907609939575,
"learning_rate": 1.819152044288992e-05,
"loss": 0.0292,
"step": 276
},
{
"epoch": 0.02216,
"grad_norm": 3.6786766052246094,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.152,
"step": 277
},
{
"epoch": 0.02224,
"grad_norm": 0.9888241291046143,
"learning_rate": 1.815127795728554e-05,
"loss": 0.0747,
"step": 278
},
{
"epoch": 0.02232,
"grad_norm": 0.48108673095703125,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.0686,
"step": 279
},
{
"epoch": 0.0224,
"grad_norm": 2.169869899749756,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.0764,
"step": 280
},
{
"epoch": 0.02248,
"grad_norm": 0.9041070938110352,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.0834,
"step": 281
},
{
"epoch": 0.02256,
"grad_norm": 0.677628219127655,
"learning_rate": 1.806960312143802e-05,
"loss": 0.0745,
"step": 282
},
{
"epoch": 0.02264,
"grad_norm": 14.399443626403809,
"learning_rate": 1.804893797355914e-05,
"loss": 0.4048,
"step": 283
},
{
"epoch": 0.02272,
"grad_norm": 1.0360500812530518,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.07,
"step": 284
},
{
"epoch": 0.0228,
"grad_norm": 0.33372780680656433,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.057,
"step": 285
},
{
"epoch": 0.02288,
"grad_norm": 1.546594262123108,
"learning_rate": 1.798635510047293e-05,
"loss": 0.0826,
"step": 286
},
{
"epoch": 0.02296,
"grad_norm": 1.3639631271362305,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.0769,
"step": 287
},
{
"epoch": 0.02304,
"grad_norm": 0.16366879642009735,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.0021,
"step": 288
},
{
"epoch": 0.02312,
"grad_norm": 1.1075043678283691,
"learning_rate": 1.792289643355191e-05,
"loss": 0.0965,
"step": 289
},
{
"epoch": 0.0232,
"grad_norm": 1.3766179084777832,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.1174,
"step": 290
},
{
"epoch": 0.02328,
"grad_norm": 1.6622263193130493,
"learning_rate": 1.788010753606722e-05,
"loss": 0.13,
"step": 291
},
{
"epoch": 0.02336,
"grad_norm": 1.2144845724105835,
"learning_rate": 1.785856893175402e-05,
"loss": 0.1077,
"step": 292
},
{
"epoch": 0.02344,
"grad_norm": 3.3310115337371826,
"learning_rate": 1.78369345732584e-05,
"loss": 0.1141,
"step": 293
},
{
"epoch": 0.02352,
"grad_norm": 0.7857168316841125,
"learning_rate": 1.781520472418819e-05,
"loss": 0.0966,
"step": 294
},
{
"epoch": 0.0236,
"grad_norm": 0.9455631375312805,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.0798,
"step": 295
},
{
"epoch": 0.02368,
"grad_norm": 0.6644325852394104,
"learning_rate": 1.777145961456971e-05,
"loss": 0.0597,
"step": 296
},
{
"epoch": 0.02376,
"grad_norm": 0.7481357455253601,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.075,
"step": 297
},
{
"epoch": 0.02384,
"grad_norm": 0.9535890221595764,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.0751,
"step": 298
},
{
"epoch": 0.02392,
"grad_norm": 0.5458697080612183,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.0679,
"step": 299
},
{
"epoch": 0.024,
"grad_norm": 0.5458417534828186,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.0666,
"step": 300
},
{
"epoch": 0.02408,
"grad_norm": 0.7755869030952454,
"learning_rate": 1.766044443118978e-05,
"loss": 0.0639,
"step": 301
},
{
"epoch": 0.02416,
"grad_norm": 0.8508099913597107,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.0983,
"step": 302
},
{
"epoch": 0.02424,
"grad_norm": 0.6589767932891846,
"learning_rate": 1.761538307536737e-05,
"loss": 0.0604,
"step": 303
},
{
"epoch": 0.02432,
"grad_norm": 0.71384197473526,
"learning_rate": 1.759271307334881e-05,
"loss": 0.0627,
"step": 304
},
{
"epoch": 0.0244,
"grad_norm": 3.834758996963501,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.0605,
"step": 305
},
{
"epoch": 0.02448,
"grad_norm": 0.047622743993997574,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.0008,
"step": 306
},
{
"epoch": 0.02456,
"grad_norm": 3.8236846923828125,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.149,
"step": 307
},
{
"epoch": 0.02464,
"grad_norm": 0.055173471570014954,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.0004,
"step": 308
},
{
"epoch": 0.02472,
"grad_norm": 1.243716835975647,
"learning_rate": 1.747798090498532e-05,
"loss": 0.0866,
"step": 309
},
{
"epoch": 0.0248,
"grad_norm": 0.7280524969100952,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.0535,
"step": 310
},
{
"epoch": 0.02488,
"grad_norm": 0.8576182723045349,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.0578,
"step": 311
},
{
"epoch": 0.02496,
"grad_norm": 0.6270462274551392,
"learning_rate": 1.74080459628675e-05,
"loss": 0.0609,
"step": 312
},
{
"epoch": 0.02504,
"grad_norm": 0.14570452272891998,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.0004,
"step": 313
},
{
"epoch": 0.02512,
"grad_norm": 0.798323929309845,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.0654,
"step": 314
},
{
"epoch": 0.0252,
"grad_norm": 10.45984172821045,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.0984,
"step": 315
},
{
"epoch": 0.02528,
"grad_norm": 1.1085079908370972,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.0668,
"step": 316
},
{
"epoch": 0.02536,
"grad_norm": 0.1317863017320633,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.0005,
"step": 317
},
{
"epoch": 0.02544,
"grad_norm": 4.704370021820068,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.1252,
"step": 318
},
{
"epoch": 0.02552,
"grad_norm": 0.624153733253479,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.0832,
"step": 319
},
{
"epoch": 0.0256,
"grad_norm": 12.77319622039795,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.3454,
"step": 320
},
{
"epoch": 0.02568,
"grad_norm": 2.1664323806762695,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.0805,
"step": 321
},
{
"epoch": 0.02576,
"grad_norm": 0.9555838108062744,
"learning_rate": 1.716910607650483e-05,
"loss": 0.0643,
"step": 322
},
{
"epoch": 0.02584,
"grad_norm": 0.6166096329689026,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.0391,
"step": 323
},
{
"epoch": 0.02592,
"grad_norm": 1.0128228664398193,
"learning_rate": 1.712026045990997e-05,
"loss": 0.0599,
"step": 324
},
{
"epoch": 0.026,
"grad_norm": 0.998466432094574,
"learning_rate": 1.709570736536521e-05,
"loss": 0.0796,
"step": 325
},
{
"epoch": 0.02608,
"grad_norm": 1.2118428945541382,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.0897,
"step": 326
},
{
"epoch": 0.02616,
"grad_norm": 1.6005258560180664,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.0914,
"step": 327
},
{
"epoch": 0.02624,
"grad_norm": 0.7567417621612549,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.064,
"step": 328
},
{
"epoch": 0.02632,
"grad_norm": 5.459105014801025,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.1612,
"step": 329
},
{
"epoch": 0.0264,
"grad_norm": 0.2641073167324066,
"learning_rate": 1.697165102854565e-05,
"loss": 0.0045,
"step": 330
},
{
"epoch": 0.02648,
"grad_norm": 1.0779212713241577,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.0846,
"step": 331
},
{
"epoch": 0.02656,
"grad_norm": 0.7016695141792297,
"learning_rate": 1.692143173870407e-05,
"loss": 0.059,
"step": 332
},
{
"epoch": 0.02664,
"grad_norm": 1.086538314819336,
"learning_rate": 1.68961954373567e-05,
"loss": 0.0615,
"step": 333
},
{
"epoch": 0.02672,
"grad_norm": 0.795050323009491,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.0588,
"step": 334
},
{
"epoch": 0.0268,
"grad_norm": 7.327888011932373,
"learning_rate": 1.684547105928689e-05,
"loss": 0.0436,
"step": 335
},
{
"epoch": 0.02688,
"grad_norm": 0.7849997878074646,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.0604,
"step": 336
},
{
"epoch": 0.02696,
"grad_norm": 0.6103646159172058,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.0573,
"step": 337
},
{
"epoch": 0.02704,
"grad_norm": 0.6351644992828369,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.0603,
"step": 338
},
{
"epoch": 0.02712,
"grad_norm": 1.2495090961456299,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.1018,
"step": 339
},
{
"epoch": 0.0272,
"grad_norm": 0.9062171578407288,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.0775,
"step": 340
},
{
"epoch": 0.02728,
"grad_norm": 0.5646159052848816,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.0601,
"step": 341
},
{
"epoch": 0.02736,
"grad_norm": 0.928389310836792,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.0626,
"step": 342
},
{
"epoch": 0.02744,
"grad_norm": 0.6078507900238037,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.0065,
"step": 343
},
{
"epoch": 0.02752,
"grad_norm": 0.6178092956542969,
"learning_rate": 1.661311865323652e-05,
"loss": 0.0501,
"step": 344
},
{
"epoch": 0.0276,
"grad_norm": 2.9329590797424316,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.1576,
"step": 345
},
{
"epoch": 0.02768,
"grad_norm": 0.7040148973464966,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.0577,
"step": 346
},
{
"epoch": 0.02776,
"grad_norm": 1.2692161798477173,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.0759,
"step": 347
},
{
"epoch": 0.02784,
"grad_norm": 0.7263659834861755,
"learning_rate": 1.650774217265851e-05,
"loss": 0.0607,
"step": 348
},
{
"epoch": 0.02792,
"grad_norm": 1.9423307180404663,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.0778,
"step": 349
},
{
"epoch": 0.028,
"grad_norm": 0.7066252827644348,
"learning_rate": 1.645457687723951e-05,
"loss": 0.0661,
"step": 350
},
{
"epoch": 0.02808,
"grad_norm": 0.7689609527587891,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.0638,
"step": 351
},
{
"epoch": 0.02816,
"grad_norm": 8.175423622131348,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.6315,
"step": 352
},
{
"epoch": 0.02824,
"grad_norm": 0.41665005683898926,
"learning_rate": 1.63742398974869e-05,
"loss": 0.0502,
"step": 353
},
{
"epoch": 0.02832,
"grad_norm": 1.4227436780929565,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.0857,
"step": 354
},
{
"epoch": 0.0284,
"grad_norm": 0.8541610240936279,
"learning_rate": 1.632029302664851e-05,
"loss": 0.0591,
"step": 355
},
{
"epoch": 0.02848,
"grad_norm": 0.4520259201526642,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.0624,
"step": 356
},
{
"epoch": 0.02856,
"grad_norm": 1.347630500793457,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.099,
"step": 357
},
{
"epoch": 0.02864,
"grad_norm": 2.18406343460083,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.0276,
"step": 358
},
{
"epoch": 0.02872,
"grad_norm": 1.0680880546569824,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.0626,
"step": 359
},
{
"epoch": 0.0288,
"grad_norm": 0.8258585333824158,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.0796,
"step": 360
},
{
"epoch": 0.02888,
"grad_norm": 0.8986599445343018,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.07,
"step": 361
},
{
"epoch": 0.02896,
"grad_norm": 0.6269177794456482,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.071,
"step": 362
},
{
"epoch": 0.02904,
"grad_norm": 0.8417167067527771,
"learning_rate": 1.610145163901268e-05,
"loss": 0.0775,
"step": 363
},
{
"epoch": 0.02912,
"grad_norm": 8.24163818359375,
"learning_rate": 1.607375839723287e-05,
"loss": 0.3217,
"step": 364
},
{
"epoch": 0.0292,
"grad_norm": 0.4157542586326599,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.03,
"step": 365
},
{
"epoch": 0.02928,
"grad_norm": 0.47345107793807983,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.0516,
"step": 366
},
{
"epoch": 0.02936,
"grad_norm": 0.8233831524848938,
"learning_rate": 1.599023598515586e-05,
"loss": 0.0358,
"step": 367
},
{
"epoch": 0.02944,
"grad_norm": 0.7989028692245483,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.0823,
"step": 368
},
{
"epoch": 0.02952,
"grad_norm": 0.6928958296775818,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.0579,
"step": 369
},
{
"epoch": 0.0296,
"grad_norm": 1.2558813095092773,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.0764,
"step": 370
},
{
"epoch": 0.02968,
"grad_norm": 0.473481148481369,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0479,
"step": 371
},
{
"epoch": 0.02976,
"grad_norm": 1.4489105939865112,
"learning_rate": 1.584957674987216e-05,
"loss": 0.0939,
"step": 372
},
{
"epoch": 0.02984,
"grad_norm": 1.9322885274887085,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.0951,
"step": 373
},
{
"epoch": 0.02992,
"grad_norm": 0.7034196257591248,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.0553,
"step": 374
},
{
"epoch": 0.03,
"grad_norm": 1.2547998428344727,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.0624,
"step": 375
},
{
"epoch": 0.03008,
"grad_norm": 0.8321748971939087,
"learning_rate": 1.573576436351046e-05,
"loss": 0.0733,
"step": 376
},
{
"epoch": 0.03016,
"grad_norm": 0.622491717338562,
"learning_rate": 1.570713567684432e-05,
"loss": 0.0551,
"step": 377
},
{
"epoch": 0.03024,
"grad_norm": 0.7036299705505371,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.065,
"step": 378
},
{
"epoch": 0.03032,
"grad_norm": 1.1212679147720337,
"learning_rate": 1.564967003424938e-05,
"loss": 0.0645,
"step": 379
},
{
"epoch": 0.0304,
"grad_norm": 1.126389503479004,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.0445,
"step": 380
},
{
"epoch": 0.03048,
"grad_norm": 0.9193438291549683,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.0786,
"step": 381
},
{
"epoch": 0.03056,
"grad_norm": 0.6234734058380127,
"learning_rate": 1.556295615500305e-05,
"loss": 0.0418,
"step": 382
},
{
"epoch": 0.03064,
"grad_norm": 0.9941009879112244,
"learning_rate": 1.553391549243344e-05,
"loss": 0.0365,
"step": 383
},
{
"epoch": 0.03072,
"grad_norm": 0.7766782641410828,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.0483,
"step": 384
},
{
"epoch": 0.0308,
"grad_norm": 1.1941871643066406,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.0645,
"step": 385
},
{
"epoch": 0.03088,
"grad_norm": 1.6085561513900757,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.0756,
"step": 386
},
{
"epoch": 0.03096,
"grad_norm": 0.3958737552165985,
"learning_rate": 1.54170821028274e-05,
"loss": 0.0355,
"step": 387
},
{
"epoch": 0.03104,
"grad_norm": 0.35567548871040344,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.0526,
"step": 388
},
{
"epoch": 0.03112,
"grad_norm": 0.11569754779338837,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.0013,
"step": 389
},
{
"epoch": 0.0312,
"grad_norm": 0.19626924395561218,
"learning_rate": 1.53287627607073e-05,
"loss": 0.002,
"step": 390
},
{
"epoch": 0.03128,
"grad_norm": 0.055757779628038406,
"learning_rate": 1.529919264233205e-05,
"loss": 0.0007,
"step": 391
},
{
"epoch": 0.03136,
"grad_norm": 0.21397146582603455,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.0316,
"step": 392
},
{
"epoch": 0.03144,
"grad_norm": 1.1726216077804565,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.0588,
"step": 393
},
{
"epoch": 0.03152,
"grad_norm": 0.8038904070854187,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.043,
"step": 394
},
{
"epoch": 0.0316,
"grad_norm": 0.09592056274414062,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.0014,
"step": 395
},
{
"epoch": 0.03168,
"grad_norm": 0.6686990857124329,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.0471,
"step": 396
},
{
"epoch": 0.03176,
"grad_norm": 1.657657265663147,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.0468,
"step": 397
},
{
"epoch": 0.03184,
"grad_norm": 0.795865535736084,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.0845,
"step": 398
},
{
"epoch": 0.03192,
"grad_norm": 0.6644959449768066,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.063,
"step": 399
},
{
"epoch": 0.032,
"grad_norm": 1.2483056783676147,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.0941,
"step": 400
},
{
"epoch": 0.03208,
"grad_norm": 0.6204407215118408,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.0729,
"step": 401
},
{
"epoch": 0.03216,
"grad_norm": 19.034936904907227,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.6528,
"step": 402
},
{
"epoch": 0.03224,
"grad_norm": 6.064889907836914,
"learning_rate": 1.493941866584231e-05,
"loss": 0.1096,
"step": 403
},
{
"epoch": 0.03232,
"grad_norm": 14.452326774597168,
"learning_rate": 1.490903753615141e-05,
"loss": 0.1901,
"step": 404
},
{
"epoch": 0.0324,
"grad_norm": 0.40182772278785706,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.0494,
"step": 405
},
{
"epoch": 0.03248,
"grad_norm": 0.8973346948623657,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.08,
"step": 406
},
{
"epoch": 0.03256,
"grad_norm": 0.9069917798042297,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.0607,
"step": 407
},
{
"epoch": 0.03264,
"grad_norm": 2.5470235347747803,
"learning_rate": 1.478691857940607e-05,
"loss": 0.0687,
"step": 408
},
{
"epoch": 0.03272,
"grad_norm": 2.092479705810547,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.0499,
"step": 409
},
{
"epoch": 0.0328,
"grad_norm": 0.7505635023117065,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.0605,
"step": 410
},
{
"epoch": 0.03288,
"grad_norm": 0.550226628780365,
"learning_rate": 1.469471562785891e-05,
"loss": 0.0403,
"step": 411
},
{
"epoch": 0.03296,
"grad_norm": 0.5675300359725952,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.0717,
"step": 412
},
{
"epoch": 0.03304,
"grad_norm": 0.7369118928909302,
"learning_rate": 1.463296035119862e-05,
"loss": 0.0646,
"step": 413
},
{
"epoch": 0.03312,
"grad_norm": 0.5541167855262756,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.0603,
"step": 414
},
{
"epoch": 0.0332,
"grad_norm": 0.37676891684532166,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.0442,
"step": 415
},
{
"epoch": 0.03328,
"grad_norm": 0.6099328994750977,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.0621,
"step": 416
},
{
"epoch": 0.03336,
"grad_norm": 3.695516347885132,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.1257,
"step": 417
},
{
"epoch": 0.03344,
"grad_norm": 1.0219265222549438,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.0737,
"step": 418
},
{
"epoch": 0.03352,
"grad_norm": 0.6232692003250122,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.071,
"step": 419
},
{
"epoch": 0.0336,
"grad_norm": 1.2602683305740356,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.067,
"step": 420
},
{
"epoch": 0.03368,
"grad_norm": 0.08829483389854431,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.0002,
"step": 421
},
{
"epoch": 0.03376,
"grad_norm": 0.8766517043113708,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.0571,
"step": 422
},
{
"epoch": 0.03384,
"grad_norm": 0.7558977603912354,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.0472,
"step": 423
},
{
"epoch": 0.03392,
"grad_norm": 1.484148621559143,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.0719,
"step": 424
},
{
"epoch": 0.034,
"grad_norm": 0.8700476288795471,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.0576,
"step": 425
},
{
"epoch": 0.03408,
"grad_norm": 4.496814727783203,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.0607,
"step": 426
},
{
"epoch": 0.03416,
"grad_norm": 3.3856918811798096,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.0762,
"step": 427
},
{
"epoch": 0.03424,
"grad_norm": 1.8766549825668335,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.0748,
"step": 428
},
{
"epoch": 0.03432,
"grad_norm": 1.242540717124939,
"learning_rate": 1.413104429824542e-05,
"loss": 0.0575,
"step": 429
},
{
"epoch": 0.0344,
"grad_norm": 0.9104759097099304,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.0721,
"step": 430
},
{
"epoch": 0.03448,
"grad_norm": 0.8081266283988953,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.0531,
"step": 431
},
{
"epoch": 0.03456,
"grad_norm": 0.6634921431541443,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.052,
"step": 432
},
{
"epoch": 0.03464,
"grad_norm": 0.5350862741470337,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.0538,
"step": 433
},
{
"epoch": 0.03472,
"grad_norm": 0.03973368927836418,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.0003,
"step": 434
},
{
"epoch": 0.0348,
"grad_norm": 0.8390694856643677,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.0468,
"step": 435
},
{
"epoch": 0.03488,
"grad_norm": 1.5010918378829956,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.0693,
"step": 436
},
{
"epoch": 0.03496,
"grad_norm": 0.040648333728313446,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.0004,
"step": 437
},
{
"epoch": 0.03504,
"grad_norm": 0.7258415222167969,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.0701,
"step": 438
},
{
"epoch": 0.03512,
"grad_norm": 0.37818005681037903,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.0363,
"step": 439
},
{
"epoch": 0.0352,
"grad_norm": 0.6038250923156738,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.0457,
"step": 440
},
{
"epoch": 0.03528,
"grad_norm": 0.030539924278855324,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.001,
"step": 441
},
{
"epoch": 0.03536,
"grad_norm": 0.9934044480323792,
"learning_rate": 1.371367835550235e-05,
"loss": 0.0885,
"step": 442
},
{
"epoch": 0.03544,
"grad_norm": 1.577244758605957,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.0575,
"step": 443
},
{
"epoch": 0.03552,
"grad_norm": 0.6882873773574829,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.0428,
"step": 444
},
{
"epoch": 0.0356,
"grad_norm": 0.03521396219730377,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.0002,
"step": 445
},
{
"epoch": 0.03568,
"grad_norm": 0.347547709941864,
"learning_rate": 1.3583679495453e-05,
"loss": 0.0424,
"step": 446
},
{
"epoch": 0.03576,
"grad_norm": 0.4582594633102417,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.0637,
"step": 447
},
{
"epoch": 0.03584,
"grad_norm": 0.5355582237243652,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.0432,
"step": 448
},
{
"epoch": 0.03592,
"grad_norm": 1.4089454412460327,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.06,
"step": 449
},
{
"epoch": 0.036,
"grad_norm": 0.8048257231712341,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.0712,
"step": 450
},
{
"epoch": 0.03608,
"grad_norm": 0.4808093011379242,
"learning_rate": 1.342020143325669e-05,
"loss": 0.0516,
"step": 451
},
{
"epoch": 0.03616,
"grad_norm": 0.8531983494758606,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.0514,
"step": 452
},
{
"epoch": 0.03624,
"grad_norm": 1.0991699695587158,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.1083,
"step": 453
},
{
"epoch": 0.03632,
"grad_norm": 3.1505355834960938,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.113,
"step": 454
},
{
"epoch": 0.0364,
"grad_norm": 0.42735937237739563,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.0417,
"step": 455
},
{
"epoch": 0.03648,
"grad_norm": 0.4284392297267914,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.0474,
"step": 456
},
{
"epoch": 0.03656,
"grad_norm": 0.5359640717506409,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.0495,
"step": 457
},
{
"epoch": 0.03664,
"grad_norm": 0.9331356883049011,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.0093,
"step": 458
},
{
"epoch": 0.03672,
"grad_norm": 1.5975041389465332,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.0699,
"step": 459
},
{
"epoch": 0.0368,
"grad_norm": 0.5402669310569763,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.0448,
"step": 460
},
{
"epoch": 0.03688,
"grad_norm": 1.3051782846450806,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.0456,
"step": 461
},
{
"epoch": 0.03696,
"grad_norm": 0.711952269077301,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.0624,
"step": 462
},
{
"epoch": 0.03704,
"grad_norm": 1.403478741645813,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.0578,
"step": 463
},
{
"epoch": 0.03712,
"grad_norm": 0.979668915271759,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.0494,
"step": 464
},
{
"epoch": 0.0372,
"grad_norm": 0.4025086462497711,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.0433,
"step": 465
},
{
"epoch": 0.03728,
"grad_norm": 0.17745329439640045,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.0005,
"step": 466
},
{
"epoch": 0.03736,
"grad_norm": 0.6765649318695068,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.0517,
"step": 467
},
{
"epoch": 0.03744,
"grad_norm": 0.7794192433357239,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.0417,
"step": 468
},
{
"epoch": 0.03752,
"grad_norm": 0.6034989953041077,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.0447,
"step": 469
},
{
"epoch": 0.0376,
"grad_norm": 0.04134514927864075,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.0008,
"step": 470
},
{
"epoch": 0.03768,
"grad_norm": 1.265798568725586,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.0676,
"step": 471
},
{
"epoch": 0.03776,
"grad_norm": 0.6860651969909668,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.0353,
"step": 472
},
{
"epoch": 0.03784,
"grad_norm": 1.6248447895050049,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.058,
"step": 473
},
{
"epoch": 0.03792,
"grad_norm": 0.4757497310638428,
"learning_rate": 1.265556117486809e-05,
"loss": 0.028,
"step": 474
},
{
"epoch": 0.038,
"grad_norm": 0.26512065529823303,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.0343,
"step": 475
},
{
"epoch": 0.03808,
"grad_norm": 0.3877088725566864,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.0439,
"step": 476
},
{
"epoch": 0.03816,
"grad_norm": 0.6541510820388794,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.0662,
"step": 477
},
{
"epoch": 0.03824,
"grad_norm": 0.6822430491447449,
"learning_rate": 1.252069358243114e-05,
"loss": 0.0476,
"step": 478
},
{
"epoch": 0.03832,
"grad_norm": 0.9895480275154114,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.0677,
"step": 479
},
{
"epoch": 0.0384,
"grad_norm": 0.26247987151145935,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.0328,
"step": 480
},
{
"epoch": 0.03848,
"grad_norm": 0.03149699792265892,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.0005,
"step": 481
},
{
"epoch": 0.03856,
"grad_norm": 0.3521241545677185,
"learning_rate": 1.238533457578581e-05,
"loss": 0.0443,
"step": 482
},
{
"epoch": 0.03864,
"grad_norm": 27.3525390625,
"learning_rate": 1.23514211310259e-05,
"loss": 0.2031,
"step": 483
},
{
"epoch": 0.03872,
"grad_norm": 0.9070781469345093,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.052,
"step": 484
},
{
"epoch": 0.0388,
"grad_norm": 1.629557490348816,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.0713,
"step": 485
},
{
"epoch": 0.03888,
"grad_norm": 1.0000633001327515,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.0563,
"step": 486
},
{
"epoch": 0.03896,
"grad_norm": 0.5456064939498901,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.0509,
"step": 487
},
{
"epoch": 0.03904,
"grad_norm": 1.167680025100708,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.0665,
"step": 488
},
{
"epoch": 0.03912,
"grad_norm": 0.8050796985626221,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.0389,
"step": 489
},
{
"epoch": 0.0392,
"grad_norm": 0.41436824202537537,
"learning_rate": 1.211324796455389e-05,
"loss": 0.034,
"step": 490
},
{
"epoch": 0.03928,
"grad_norm": 0.5509086847305298,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0189,
"step": 491
},
{
"epoch": 0.03936,
"grad_norm": 1.0947761535644531,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.0531,
"step": 492
},
{
"epoch": 0.03944,
"grad_norm": 0.36645254492759705,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.0456,
"step": 493
},
{
"epoch": 0.03952,
"grad_norm": 0.45027920603752136,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.047,
"step": 494
},
{
"epoch": 0.0396,
"grad_norm": 1.8112543821334839,
"learning_rate": 1.194234351219972e-05,
"loss": 0.0743,
"step": 495
},
{
"epoch": 0.03968,
"grad_norm": 0.43417835235595703,
"learning_rate": 1.190808995376545e-05,
"loss": 0.0514,
"step": 496
},
{
"epoch": 0.03976,
"grad_norm": 0.6519817113876343,
"learning_rate": 1.187381314585725e-05,
"loss": 0.0713,
"step": 497
},
{
"epoch": 0.03984,
"grad_norm": 0.49262917041778564,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.0458,
"step": 498
},
{
"epoch": 0.03992,
"grad_norm": 0.6282567977905273,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.0304,
"step": 499
},
{
"epoch": 0.04,
"grad_norm": 1.062343716621399,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.0526,
"step": 500
},
{
"epoch": 0.04008,
"grad_norm": 0.6629893779754639,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.0511,
"step": 501
},
{
"epoch": 0.04016,
"grad_norm": 0.29980793595314026,
"learning_rate": 1.1702094991660326e-05,
"loss": 0.0317,
"step": 502
},
{
"epoch": 0.04024,
"grad_norm": 0.6524478793144226,
"learning_rate": 1.1667687467161025e-05,
"loss": 0.0534,
"step": 503
},
{
"epoch": 0.04032,
"grad_norm": 0.9067026972770691,
"learning_rate": 1.1633259622416224e-05,
"loss": 0.0542,
"step": 504
},
{
"epoch": 0.0404,
"grad_norm": 0.544202983379364,
"learning_rate": 1.159881187691835e-05,
"loss": 0.0537,
"step": 505
},
{
"epoch": 0.04048,
"grad_norm": 1.1514512300491333,
"learning_rate": 1.156434465040231e-05,
"loss": 0.0844,
"step": 506
},
{
"epoch": 0.04056,
"grad_norm": 0.526523768901825,
"learning_rate": 1.1529858362840383e-05,
"loss": 0.0305,
"step": 507
},
{
"epoch": 0.04064,
"grad_norm": 10.12170696258545,
"learning_rate": 1.1495353434437098e-05,
"loss": 0.2218,
"step": 508
},
{
"epoch": 0.04072,
"grad_norm": 0.907473087310791,
"learning_rate": 1.1460830285624119e-05,
"loss": 0.037,
"step": 509
},
{
"epoch": 0.0408,
"grad_norm": 0.7351361513137817,
"learning_rate": 1.1426289337055119e-05,
"loss": 0.049,
"step": 510
},
{
"epoch": 0.04088,
"grad_norm": 0.41098979115486145,
"learning_rate": 1.1391731009600655e-05,
"loss": 0.0429,
"step": 511
},
{
"epoch": 0.04096,
"grad_norm": 0.5460197925567627,
"learning_rate": 1.1357155724343046e-05,
"loss": 0.0423,
"step": 512
},
{
"epoch": 0.04104,
"grad_norm": 0.35029977560043335,
"learning_rate": 1.1322563902571227e-05,
"loss": 0.0414,
"step": 513
},
{
"epoch": 0.04112,
"grad_norm": 0.7547082304954529,
"learning_rate": 1.128795596577563e-05,
"loss": 0.0648,
"step": 514
},
{
"epoch": 0.0412,
"grad_norm": 1.976358413696289,
"learning_rate": 1.1253332335643043e-05,
"loss": 0.0616,
"step": 515
},
{
"epoch": 0.04128,
"grad_norm": 0.37118279933929443,
"learning_rate": 1.1218693434051475e-05,
"loss": 0.0397,
"step": 516
},
{
"epoch": 0.04136,
"grad_norm": 0.03787941485643387,
"learning_rate": 1.1184039683065014e-05,
"loss": 0.0007,
"step": 517
},
{
"epoch": 0.04144,
"grad_norm": 0.02803988754749298,
"learning_rate": 1.1149371504928667e-05,
"loss": 0.0006,
"step": 518
},
{
"epoch": 0.04152,
"grad_norm": 1.0917259454727173,
"learning_rate": 1.1114689322063255e-05,
"loss": 0.0701,
"step": 519
},
{
"epoch": 0.0416,
"grad_norm": 0.028895191848278046,
"learning_rate": 1.1079993557060228e-05,
"loss": 0.0005,
"step": 520
},
{
"epoch": 0.04168,
"grad_norm": 0.8043434619903564,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.0832,
"step": 521
},
{
"epoch": 0.04176,
"grad_norm": 0.2628927230834961,
"learning_rate": 1.1010562971829464e-05,
"loss": 0.0307,
"step": 522
},
{
"epoch": 0.04184,
"grad_norm": 0.2679750323295593,
"learning_rate": 1.0975828997591496e-05,
"loss": 0.0381,
"step": 523
},
{
"epoch": 0.04192,
"grad_norm": 0.5351852774620056,
"learning_rate": 1.0941083133185146e-05,
"loss": 0.0324,
"step": 524
},
{
"epoch": 0.042,
"grad_norm": 0.9442641735076904,
"learning_rate": 1.0906325801977804e-05,
"loss": 0.0847,
"step": 525
},
{
"epoch": 0.04208,
"grad_norm": 0.3288194239139557,
"learning_rate": 1.0871557427476585e-05,
"loss": 0.0326,
"step": 526
},
{
"epoch": 0.04216,
"grad_norm": 0.225941002368927,
"learning_rate": 1.083677843332316e-05,
"loss": 0.0304,
"step": 527
},
{
"epoch": 0.04224,
"grad_norm": 0.5145255327224731,
"learning_rate": 1.0801989243288588e-05,
"loss": 0.0349,
"step": 528
},
{
"epoch": 0.04232,
"grad_norm": 0.2666696012020111,
"learning_rate": 1.0767190281268187e-05,
"loss": 0.05,
"step": 529
},
{
"epoch": 0.0424,
"grad_norm": 0.581710934638977,
"learning_rate": 1.0732381971276318e-05,
"loss": 0.0499,
"step": 530
},
{
"epoch": 0.04248,
"grad_norm": 1.8520171642303467,
"learning_rate": 1.0697564737441254e-05,
"loss": 0.1053,
"step": 531
},
{
"epoch": 0.04256,
"grad_norm": 1.8509712219238281,
"learning_rate": 1.0662739004000005e-05,
"loss": 0.0788,
"step": 532
},
{
"epoch": 0.04264,
"grad_norm": 6.737593173980713,
"learning_rate": 1.0627905195293135e-05,
"loss": 0.1807,
"step": 533
},
{
"epoch": 0.04272,
"grad_norm": 0.4677772521972656,
"learning_rate": 1.0593063735759619e-05,
"loss": 0.0486,
"step": 534
},
{
"epoch": 0.0428,
"grad_norm": 0.3308097720146179,
"learning_rate": 1.055821504993164e-05,
"loss": 0.032,
"step": 535
},
{
"epoch": 0.04288,
"grad_norm": 0.49913105368614197,
"learning_rate": 1.0523359562429441e-05,
"loss": 0.0376,
"step": 536
},
{
"epoch": 0.04296,
"grad_norm": 0.2439991682767868,
"learning_rate": 1.0488497697956134e-05,
"loss": 0.0023,
"step": 537
},
{
"epoch": 0.04304,
"grad_norm": 0.46642744541168213,
"learning_rate": 1.0453629881292537e-05,
"loss": 0.0326,
"step": 538
},
{
"epoch": 0.04312,
"grad_norm": 0.381874680519104,
"learning_rate": 1.0418756537291996e-05,
"loss": 0.0485,
"step": 539
},
{
"epoch": 0.0432,
"grad_norm": 0.9421465396881104,
"learning_rate": 1.03838780908752e-05,
"loss": 0.0376,
"step": 540
},
{
"epoch": 0.04328,
"grad_norm": 0.4877367615699768,
"learning_rate": 1.0348994967025012e-05,
"loss": 0.0501,
"step": 541
},
{
"epoch": 0.04336,
"grad_norm": 0.022308025509119034,
"learning_rate": 1.0314107590781284e-05,
"loss": 0.0006,
"step": 542
},
{
"epoch": 0.04344,
"grad_norm": 0.7615314722061157,
"learning_rate": 1.0279216387235691e-05,
"loss": 0.0599,
"step": 543
},
{
"epoch": 0.04352,
"grad_norm": 0.41580528020858765,
"learning_rate": 1.0244321781526533e-05,
"loss": 0.0285,
"step": 544
},
{
"epoch": 0.0436,
"grad_norm": 0.974173903465271,
"learning_rate": 1.0209424198833571e-05,
"loss": 0.0775,
"step": 545
},
{
"epoch": 0.04368,
"grad_norm": 0.48370206356048584,
"learning_rate": 1.0174524064372837e-05,
"loss": 0.0555,
"step": 546
},
{
"epoch": 0.04376,
"grad_norm": 0.49840739369392395,
"learning_rate": 1.0139621803391454e-05,
"loss": 0.0464,
"step": 547
},
{
"epoch": 0.04384,
"grad_norm": 0.381291925907135,
"learning_rate": 1.010471784116246e-05,
"loss": 0.0326,
"step": 548
},
{
"epoch": 0.04392,
"grad_norm": 0.02992522530257702,
"learning_rate": 1.0069812602979617e-05,
"loss": 0.0003,
"step": 549
},
{
"epoch": 0.044,
"grad_norm": 1.5620512962341309,
"learning_rate": 1.0034906514152239e-05,
"loss": 0.0554,
"step": 550
},
{
"epoch": 0.04408,
"grad_norm": 0.10739638656377792,
"learning_rate": 1e-05,
"loss": 0.0005,
"step": 551
},
{
"epoch": 0.04416,
"grad_norm": 1.5892431735992432,
"learning_rate": 9.965093485847766e-06,
"loss": 0.0406,
"step": 552
},
{
"epoch": 0.04424,
"grad_norm": 16.732378005981445,
"learning_rate": 9.930187397020385e-06,
"loss": 0.6351,
"step": 553
},
{
"epoch": 0.04432,
"grad_norm": 0.044130854308605194,
"learning_rate": 9.895282158837545e-06,
"loss": 0.0012,
"step": 554
},
{
"epoch": 0.0444,
"grad_norm": 0.8484880924224854,
"learning_rate": 9.860378196608549e-06,
"loss": 0.0337,
"step": 555
},
{
"epoch": 0.04448,
"grad_norm": 0.34229961037635803,
"learning_rate": 9.825475935627165e-06,
"loss": 0.0378,
"step": 556
},
{
"epoch": 0.04456,
"grad_norm": 0.3770654499530792,
"learning_rate": 9.790575801166432e-06,
"loss": 0.0441,
"step": 557
},
{
"epoch": 0.04464,
"grad_norm": 0.5559030175209045,
"learning_rate": 9.75567821847347e-06,
"loss": 0.0542,
"step": 558
},
{
"epoch": 0.04472,
"grad_norm": 0.7541608810424805,
"learning_rate": 9.720783612764314e-06,
"loss": 0.0642,
"step": 559
},
{
"epoch": 0.0448,
"grad_norm": 0.4217135012149811,
"learning_rate": 9.685892409218718e-06,
"loss": 0.0357,
"step": 560
},
{
"epoch": 0.04488,
"grad_norm": 2.5795137882232666,
"learning_rate": 9.651005032974994e-06,
"loss": 0.0871,
"step": 561
},
{
"epoch": 0.04496,
"grad_norm": 2.6235694885253906,
"learning_rate": 9.616121909124801e-06,
"loss": 0.0775,
"step": 562
},
{
"epoch": 0.04504,
"grad_norm": 1.4136192798614502,
"learning_rate": 9.581243462708007e-06,
"loss": 0.0647,
"step": 563
},
{
"epoch": 0.04512,
"grad_norm": 0.4257340431213379,
"learning_rate": 9.546370118707463e-06,
"loss": 0.0439,
"step": 564
},
{
"epoch": 0.0452,
"grad_norm": 1.068382978439331,
"learning_rate": 9.511502302043867e-06,
"loss": 0.0363,
"step": 565
},
{
"epoch": 0.04528,
"grad_norm": 1.7305870056152344,
"learning_rate": 9.476640437570562e-06,
"loss": 0.0548,
"step": 566
},
{
"epoch": 0.04536,
"grad_norm": 0.6882795095443726,
"learning_rate": 9.441784950068362e-06,
"loss": 0.0484,
"step": 567
},
{
"epoch": 0.04544,
"grad_norm": 0.48330432176589966,
"learning_rate": 9.406936264240386e-06,
"loss": 0.0511,
"step": 568
},
{
"epoch": 0.04552,
"grad_norm": 1.1300121545791626,
"learning_rate": 9.372094804706867e-06,
"loss": 0.0362,
"step": 569
},
{
"epoch": 0.0456,
"grad_norm": 0.035712964832782745,
"learning_rate": 9.337260996000002e-06,
"loss": 0.0,
"step": 570
},
{
"epoch": 0.04568,
"grad_norm": 0.28784191608428955,
"learning_rate": 9.302435262558748e-06,
"loss": 0.0311,
"step": 571
},
{
"epoch": 0.04576,
"grad_norm": 0.9785875678062439,
"learning_rate": 9.267618028723687e-06,
"loss": 0.0906,
"step": 572
},
{
"epoch": 0.04584,
"grad_norm": 0.4081527888774872,
"learning_rate": 9.232809718731815e-06,
"loss": 0.036,
"step": 573
},
{
"epoch": 0.04592,
"grad_norm": 21.377731323242188,
"learning_rate": 9.198010756711413e-06,
"loss": 0.1379,
"step": 574
},
{
"epoch": 0.046,
"grad_norm": 0.9246197938919067,
"learning_rate": 9.163221566676847e-06,
"loss": 0.0541,
"step": 575
},
{
"epoch": 0.04608,
"grad_norm": 0.33143576979637146,
"learning_rate": 9.128442572523418e-06,
"loss": 0.0494,
"step": 576
},
{
"epoch": 0.04616,
"grad_norm": 2.4649391174316406,
"learning_rate": 9.093674198022201e-06,
"loss": 0.1018,
"step": 577
},
{
"epoch": 0.04624,
"grad_norm": 0.2764960825443268,
"learning_rate": 9.058916866814857e-06,
"loss": 0.0412,
"step": 578
},
{
"epoch": 0.04632,
"grad_norm": 0.4980447590351105,
"learning_rate": 9.024171002408507e-06,
"loss": 0.0753,
"step": 579
},
{
"epoch": 0.0464,
"grad_norm": 0.3540217876434326,
"learning_rate": 8.989437028170537e-06,
"loss": 0.0551,
"step": 580
},
{
"epoch": 0.04648,
"grad_norm": 1.0394651889801025,
"learning_rate": 8.954715367323468e-06,
"loss": 0.0689,
"step": 581
},
{
"epoch": 0.04656,
"grad_norm": 1.372073769569397,
"learning_rate": 8.920006442939772e-06,
"loss": 0.0339,
"step": 582
},
{
"epoch": 0.04664,
"grad_norm": 0.39932531118392944,
"learning_rate": 8.885310677936746e-06,
"loss": 0.0451,
"step": 583
},
{
"epoch": 0.04672,
"grad_norm": 0.23082974553108215,
"learning_rate": 8.850628495071336e-06,
"loss": 0.0265,
"step": 584
},
{
"epoch": 0.0468,
"grad_norm": 0.625804603099823,
"learning_rate": 8.815960316934991e-06,
"loss": 0.0455,
"step": 585
},
{
"epoch": 0.04688,
"grad_norm": 0.9905107021331787,
"learning_rate": 8.781306565948528e-06,
"loss": 0.0424,
"step": 586
},
{
"epoch": 0.04696,
"grad_norm": 1.749990701675415,
"learning_rate": 8.746667664356957e-06,
"loss": 0.0765,
"step": 587
},
{
"epoch": 0.04704,
"grad_norm": 0.7228845953941345,
"learning_rate": 8.712044034224374e-06,
"loss": 0.0648,
"step": 588
},
{
"epoch": 0.04712,
"grad_norm": 0.7181887030601501,
"learning_rate": 8.677436097428775e-06,
"loss": 0.027,
"step": 589
},
{
"epoch": 0.0472,
"grad_norm": 0.46673086285591125,
"learning_rate": 8.642844275656957e-06,
"loss": 0.0264,
"step": 590
},
{
"epoch": 0.04728,
"grad_norm": 0.8278421759605408,
"learning_rate": 8.60826899039935e-06,
"loss": 0.0552,
"step": 591
},
{
"epoch": 0.04736,
"grad_norm": 0.750532329082489,
"learning_rate": 8.573710662944884e-06,
"loss": 0.0812,
"step": 592
},
{
"epoch": 0.04744,
"grad_norm": 0.6552270650863647,
"learning_rate": 8.539169714375885e-06,
"loss": 0.0403,
"step": 593
},
{
"epoch": 0.04752,
"grad_norm": 0.559214174747467,
"learning_rate": 8.504646565562907e-06,
"loss": 0.0388,
"step": 594
},
{
"epoch": 0.0476,
"grad_norm": 0.15761122107505798,
"learning_rate": 8.47014163715962e-06,
"loss": 0.0223,
"step": 595
},
{
"epoch": 0.04768,
"grad_norm": 0.9963493943214417,
"learning_rate": 8.43565534959769e-06,
"loss": 0.0485,
"step": 596
},
{
"epoch": 0.04776,
"grad_norm": 1.2800792455673218,
"learning_rate": 8.401188123081653e-06,
"loss": 0.0587,
"step": 597
},
{
"epoch": 0.04784,
"grad_norm": 0.042244184762239456,
"learning_rate": 8.366740377583781e-06,
"loss": 0.0001,
"step": 598
},
{
"epoch": 0.04792,
"grad_norm": 0.5075722336769104,
"learning_rate": 8.332312532838978e-06,
"loss": 0.0536,
"step": 599
},
{
"epoch": 0.048,
"grad_norm": 0.05619000270962715,
"learning_rate": 8.297905008339677e-06,
"loss": 0.0007,
"step": 600
},
{
"epoch": 0.04808,
"grad_norm": 0.332971453666687,
"learning_rate": 8.263518223330698e-06,
"loss": 0.0454,
"step": 601
},
{
"epoch": 0.04816,
"grad_norm": 0.634878396987915,
"learning_rate": 8.22915259680417e-06,
"loss": 0.0181,
"step": 602
},
{
"epoch": 0.04824,
"grad_norm": 0.4374867081642151,
"learning_rate": 8.194808547494401e-06,
"loss": 0.0472,
"step": 603
},
{
"epoch": 0.04832,
"grad_norm": 0.6745132207870483,
"learning_rate": 8.1604864938728e-06,
"loss": 0.0513,
"step": 604
},
{
"epoch": 0.0484,
"grad_norm": 0.13090407848358154,
"learning_rate": 8.126186854142752e-06,
"loss": 0.0022,
"step": 605
},
{
"epoch": 0.04848,
"grad_norm": 0.31579476594924927,
"learning_rate": 8.091910046234552e-06,
"loss": 0.039,
"step": 606
},
{
"epoch": 0.04856,
"grad_norm": 0.44567400217056274,
"learning_rate": 8.057656487800283e-06,
"loss": 0.0024,
"step": 607
},
{
"epoch": 0.04864,
"grad_norm": 0.27829158306121826,
"learning_rate": 8.023426596208739e-06,
"loss": 0.0399,
"step": 608
},
{
"epoch": 0.04872,
"grad_norm": 1.409342646598816,
"learning_rate": 7.989220788540356e-06,
"loss": 0.0649,
"step": 609
},
{
"epoch": 0.0488,
"grad_norm": 0.023370087146759033,
"learning_rate": 7.955039481582098e-06,
"loss": 0.0005,
"step": 610
},
{
"epoch": 0.04888,
"grad_norm": 0.5067552924156189,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0521,
"step": 611
},
{
"epoch": 0.04896,
"grad_norm": 0.6468042135238647,
"learning_rate": 7.886752035446116e-06,
"loss": 0.046,
"step": 612
},
{
"epoch": 0.04904,
"grad_norm": 0.2538033127784729,
"learning_rate": 7.852646728329368e-06,
"loss": 0.0392,
"step": 613
},
{
"epoch": 0.04912,
"grad_norm": 0.6498031616210938,
"learning_rate": 7.818567586034578e-06,
"loss": 0.0436,
"step": 614
},
{
"epoch": 0.0492,
"grad_norm": 0.6893162727355957,
"learning_rate": 7.784515023805328e-06,
"loss": 0.0561,
"step": 615
},
{
"epoch": 0.04928,
"grad_norm": 0.27984878420829773,
"learning_rate": 7.750489456561351e-06,
"loss": 0.0461,
"step": 616
},
{
"epoch": 0.04936,
"grad_norm": 0.421854168176651,
"learning_rate": 7.716491298893443e-06,
"loss": 0.046,
"step": 617
},
{
"epoch": 0.04944,
"grad_norm": 0.22991496324539185,
"learning_rate": 7.68252096505843e-06,
"loss": 0.0212,
"step": 618
},
{
"epoch": 0.04952,
"grad_norm": 0.4577009677886963,
"learning_rate": 7.6485788689741e-06,
"loss": 0.0436,
"step": 619
},
{
"epoch": 0.0496,
"grad_norm": 0.33450445532798767,
"learning_rate": 7.6146654242141935e-06,
"loss": 0.0388,
"step": 620
},
{
"epoch": 0.04968,
"grad_norm": 0.049823690205812454,
"learning_rate": 7.580781044003324e-06,
"loss": 0.0012,
"step": 621
},
{
"epoch": 0.04976,
"grad_norm": 0.5863931179046631,
"learning_rate": 7.546926141211975e-06,
"loss": 0.0527,
"step": 622
},
{
"epoch": 0.04984,
"grad_norm": 0.4893271028995514,
"learning_rate": 7.513101128351454e-06,
"loss": 0.0328,
"step": 623
},
{
"epoch": 0.04992,
"grad_norm": 1.0855265855789185,
"learning_rate": 7.4793064175688635e-06,
"loss": 0.0566,
"step": 624
},
{
"epoch": 0.05,
"grad_norm": 2.4685475826263428,
"learning_rate": 7.445542420642097e-06,
"loss": 0.1031,
"step": 625
},
{
"epoch": 0.05008,
"grad_norm": 0.15916621685028076,
"learning_rate": 7.411809548974792e-06,
"loss": 0.0369,
"step": 626
},
{
"epoch": 0.05016,
"grad_norm": 3.4022703170776367,
"learning_rate": 7.378108213591355e-06,
"loss": 0.1179,
"step": 627
},
{
"epoch": 0.05024,
"grad_norm": 0.8043720126152039,
"learning_rate": 7.344438825131912e-06,
"loss": 0.0484,
"step": 628
},
{
"epoch": 0.05032,
"grad_norm": 0.5012772083282471,
"learning_rate": 7.310801793847344e-06,
"loss": 0.0668,
"step": 629
},
{
"epoch": 0.0504,
"grad_norm": 0.3096311390399933,
"learning_rate": 7.277197529594257e-06,
"loss": 0.0396,
"step": 630
},
{
"epoch": 0.05048,
"grad_norm": 0.36351093649864197,
"learning_rate": 7.243626441830009e-06,
"loss": 0.0524,
"step": 631
},
{
"epoch": 0.05056,
"grad_norm": 0.04093038663268089,
"learning_rate": 7.210088939607709e-06,
"loss": 0.0003,
"step": 632
},
{
"epoch": 0.05064,
"grad_norm": 0.3755309283733368,
"learning_rate": 7.176585431571235e-06,
"loss": 0.0361,
"step": 633
},
{
"epoch": 0.05072,
"grad_norm": 2.5055294036865234,
"learning_rate": 7.143116325950266e-06,
"loss": 0.0807,
"step": 634
},
{
"epoch": 0.0508,
"grad_norm": 0.4387355446815491,
"learning_rate": 7.109682030555283e-06,
"loss": 0.0434,
"step": 635
},
{
"epoch": 0.05088,
"grad_norm": 0.6049374938011169,
"learning_rate": 7.076282952772634e-06,
"loss": 0.0456,
"step": 636
},
{
"epoch": 0.05096,
"grad_norm": 1.6827545166015625,
"learning_rate": 7.042919499559538e-06,
"loss": 0.0553,
"step": 637
},
{
"epoch": 0.05104,
"grad_norm": 0.6997621059417725,
"learning_rate": 7.009592077439135e-06,
"loss": 0.0574,
"step": 638
},
{
"epoch": 0.05112,
"grad_norm": 13.46363353729248,
"learning_rate": 6.976301092495556e-06,
"loss": 0.0563,
"step": 639
},
{
"epoch": 0.0512,
"grad_norm": 1.0819239616394043,
"learning_rate": 6.943046950368944e-06,
"loss": 0.0429,
"step": 640
},
{
"epoch": 0.05128,
"grad_norm": 2.895540475845337,
"learning_rate": 6.909830056250527e-06,
"loss": 0.1025,
"step": 641
},
{
"epoch": 0.05136,
"grad_norm": 0.37224623560905457,
"learning_rate": 6.876650814877675e-06,
"loss": 0.036,
"step": 642
},
{
"epoch": 0.05144,
"grad_norm": 0.773871898651123,
"learning_rate": 6.843509630528977e-06,
"loss": 0.0545,
"step": 643
},
{
"epoch": 0.05152,
"grad_norm": 1.3716028928756714,
"learning_rate": 6.8104069070193e-06,
"loss": 0.0765,
"step": 644
},
{
"epoch": 0.0516,
"grad_norm": 0.2551344633102417,
"learning_rate": 6.777343047694891e-06,
"loss": 0.0319,
"step": 645
},
{
"epoch": 0.05168,
"grad_norm": 0.04652904346585274,
"learning_rate": 6.744318455428436e-06,
"loss": 0.0007,
"step": 646
},
{
"epoch": 0.05176,
"grad_norm": 0.5680582523345947,
"learning_rate": 6.711333532614168e-06,
"loss": 0.0555,
"step": 647
},
{
"epoch": 0.05184,
"grad_norm": 1.0083003044128418,
"learning_rate": 6.67838868116297e-06,
"loss": 0.0723,
"step": 648
},
{
"epoch": 0.05192,
"grad_norm": 0.02310173772275448,
"learning_rate": 6.645484302497452e-06,
"loss": 0.0008,
"step": 649
},
{
"epoch": 0.052,
"grad_norm": 0.19674283266067505,
"learning_rate": 6.612620797547087e-06,
"loss": 0.0294,
"step": 650
},
{
"epoch": 0.05208,
"grad_norm": 0.48939263820648193,
"learning_rate": 6.579798566743314e-06,
"loss": 0.0427,
"step": 651
},
{
"epoch": 0.05216,
"grad_norm": 0.38048484921455383,
"learning_rate": 6.547018010014654e-06,
"loss": 0.0443,
"step": 652
},
{
"epoch": 0.05224,
"grad_norm": 0.0410778783261776,
"learning_rate": 6.5142795267818505e-06,
"loss": 0.0,
"step": 653
},
{
"epoch": 0.05232,
"grad_norm": 0.3994109630584717,
"learning_rate": 6.481583515952983e-06,
"loss": 0.005,
"step": 654
},
{
"epoch": 0.0524,
"grad_norm": 0.3327992558479309,
"learning_rate": 6.448930375918632e-06,
"loss": 0.0262,
"step": 655
},
{
"epoch": 0.05248,
"grad_norm": 2.0741465091705322,
"learning_rate": 6.4163205045469975e-06,
"loss": 0.0577,
"step": 656
},
{
"epoch": 0.05256,
"grad_norm": 0.3551333248615265,
"learning_rate": 6.383754299179079e-06,
"loss": 0.0424,
"step": 657
},
{
"epoch": 0.05264,
"grad_norm": 0.6887756586074829,
"learning_rate": 6.351232156623803e-06,
"loss": 0.0633,
"step": 658
},
{
"epoch": 0.05272,
"grad_norm": 1.041481375694275,
"learning_rate": 6.318754473153221e-06,
"loss": 0.0525,
"step": 659
},
{
"epoch": 0.0528,
"grad_norm": 0.5626590251922607,
"learning_rate": 6.286321644497655e-06,
"loss": 0.06,
"step": 660
},
{
"epoch": 0.05288,
"grad_norm": 0.40421783924102783,
"learning_rate": 6.25393406584088e-06,
"loss": 0.0477,
"step": 661
},
{
"epoch": 0.05296,
"grad_norm": 0.4572801887989044,
"learning_rate": 6.22159213181533e-06,
"loss": 0.0492,
"step": 662
},
{
"epoch": 0.05304,
"grad_norm": 1.1299611330032349,
"learning_rate": 6.18929623649726e-06,
"loss": 0.0528,
"step": 663
},
{
"epoch": 0.05312,
"grad_norm": 0.027984783053398132,
"learning_rate": 6.157046773401964e-06,
"loss": 0.0007,
"step": 664
},
{
"epoch": 0.0532,
"grad_norm": 0.38015949726104736,
"learning_rate": 6.124844135478971e-06,
"loss": 0.0318,
"step": 665
},
{
"epoch": 0.05328,
"grad_norm": 0.5296044945716858,
"learning_rate": 6.092688715107265e-06,
"loss": 0.0409,
"step": 666
},
{
"epoch": 0.05336,
"grad_norm": 1.5563422441482544,
"learning_rate": 6.06058090409049e-06,
"loss": 0.0512,
"step": 667
},
{
"epoch": 0.05344,
"grad_norm": 1.0861858129501343,
"learning_rate": 6.028521093652195e-06,
"loss": 0.0964,
"step": 668
},
{
"epoch": 0.05352,
"grad_norm": 0.1879701316356659,
"learning_rate": 5.996509674431053e-06,
"loss": 0.0238,
"step": 669
},
{
"epoch": 0.0536,
"grad_norm": 0.36316928267478943,
"learning_rate": 5.9645470364761e-06,
"loss": 0.0549,
"step": 670
},
{
"epoch": 0.05368,
"grad_norm": 0.4537529945373535,
"learning_rate": 5.932633569242e-06,
"loss": 0.0464,
"step": 671
},
{
"epoch": 0.05376,
"grad_norm": 1.578576683998108,
"learning_rate": 5.900769661584273e-06,
"loss": 0.0562,
"step": 672
},
{
"epoch": 0.05384,
"grad_norm": 0.44255322217941284,
"learning_rate": 5.868955701754584e-06,
"loss": 0.0211,
"step": 673
},
{
"epoch": 0.05392,
"grad_norm": 0.7401270270347595,
"learning_rate": 5.83719207739599e-06,
"loss": 0.0494,
"step": 674
},
{
"epoch": 0.054,
"grad_norm": 0.9426566958427429,
"learning_rate": 5.8054791755382286e-06,
"loss": 0.0506,
"step": 675
},
{
"epoch": 0.05408,
"grad_norm": 0.48094046115875244,
"learning_rate": 5.773817382593008e-06,
"loss": 0.0153,
"step": 676
},
{
"epoch": 0.05416,
"grad_norm": 0.5730823278427124,
"learning_rate": 5.742207084349274e-06,
"loss": 0.0552,
"step": 677
},
{
"epoch": 0.05424,
"grad_norm": 0.18017062544822693,
"learning_rate": 5.710648665968543e-06,
"loss": 0.0329,
"step": 678
},
{
"epoch": 0.05432,
"grad_norm": 1.001624345779419,
"learning_rate": 5.679142511980176e-06,
"loss": 0.0423,
"step": 679
},
{
"epoch": 0.0544,
"grad_norm": 0.2957208454608917,
"learning_rate": 5.647689006276727e-06,
"loss": 0.0466,
"step": 680
},
{
"epoch": 0.05448,
"grad_norm": 0.8194050192832947,
"learning_rate": 5.616288532109225e-06,
"loss": 0.0466,
"step": 681
},
{
"epoch": 0.05456,
"grad_norm": 0.6524631381034851,
"learning_rate": 5.584941472082549e-06,
"loss": 0.0387,
"step": 682
},
{
"epoch": 0.05464,
"grad_norm": 0.7064331769943237,
"learning_rate": 5.553648208150728e-06,
"loss": 0.0612,
"step": 683
},
{
"epoch": 0.05472,
"grad_norm": 1.0523494482040405,
"learning_rate": 5.522409121612304e-06,
"loss": 0.0571,
"step": 684
},
{
"epoch": 0.0548,
"grad_norm": 0.635187566280365,
"learning_rate": 5.491224593105695e-06,
"loss": 0.0422,
"step": 685
},
{
"epoch": 0.05488,
"grad_norm": 0.2306356132030487,
"learning_rate": 5.460095002604533e-06,
"loss": 0.0471,
"step": 686
},
{
"epoch": 0.05496,
"grad_norm": 0.7530050873756409,
"learning_rate": 5.429020729413062e-06,
"loss": 0.0488,
"step": 687
},
{
"epoch": 0.05504,
"grad_norm": 0.44951504468917847,
"learning_rate": 5.398002152161484e-06,
"loss": 0.04,
"step": 688
},
{
"epoch": 0.05512,
"grad_norm": 2.475050926208496,
"learning_rate": 5.367039648801386e-06,
"loss": 0.0746,
"step": 689
},
{
"epoch": 0.0552,
"grad_norm": 0.3671523332595825,
"learning_rate": 5.336133596601089e-06,
"loss": 0.0188,
"step": 690
},
{
"epoch": 0.05528,
"grad_norm": 0.6504513621330261,
"learning_rate": 5.305284372141095e-06,
"loss": 0.0402,
"step": 691
},
{
"epoch": 0.05536,
"grad_norm": 0.3720727860927582,
"learning_rate": 5.274492351309462e-06,
"loss": 0.0343,
"step": 692
},
{
"epoch": 0.05544,
"grad_norm": 0.6426114439964294,
"learning_rate": 5.243757909297247e-06,
"loss": 0.0335,
"step": 693
},
{
"epoch": 0.05552,
"grad_norm": 1.450393557548523,
"learning_rate": 5.213081420593933e-06,
"loss": 0.0668,
"step": 694
},
{
"epoch": 0.0556,
"grad_norm": 1.0773845911026,
"learning_rate": 5.1824632589828465e-06,
"loss": 0.0469,
"step": 695
},
{
"epoch": 0.05568,
"grad_norm": 0.9230513572692871,
"learning_rate": 5.151903797536631e-06,
"loss": 0.0505,
"step": 696
},
{
"epoch": 0.05576,
"grad_norm": 0.29917237162590027,
"learning_rate": 5.121403408612672e-06,
"loss": 0.05,
"step": 697
},
{
"epoch": 0.05584,
"grad_norm": 0.7286171317100525,
"learning_rate": 5.090962463848592e-06,
"loss": 0.0481,
"step": 698
},
{
"epoch": 0.05592,
"grad_norm": 0.40322861075401306,
"learning_rate": 5.060581334157693e-06,
"loss": 0.0415,
"step": 699
},
{
"epoch": 0.056,
"grad_norm": 0.38930150866508484,
"learning_rate": 5.030260389724447e-06,
"loss": 0.0452,
"step": 700
},
{
"epoch": 0.05608,
"grad_norm": 0.6384855508804321,
"learning_rate": 5.000000000000003e-06,
"loss": 0.0528,
"step": 701
},
{
"epoch": 0.05616,
"grad_norm": 0.6833522915840149,
"learning_rate": 4.96980053369765e-06,
"loss": 0.0469,
"step": 702
},
{
"epoch": 0.05624,
"grad_norm": 0.4922811985015869,
"learning_rate": 4.939662358788364e-06,
"loss": 0.0133,
"step": 703
},
{
"epoch": 0.05632,
"grad_norm": 12.706682205200195,
"learning_rate": 4.909585842496287e-06,
"loss": 0.7731,
"step": 704
},
{
"epoch": 0.0564,
"grad_norm": 0.3512198030948639,
"learning_rate": 4.879571351294287e-06,
"loss": 0.0485,
"step": 705
},
{
"epoch": 0.05648,
"grad_norm": 1.5541528463363647,
"learning_rate": 4.849619250899458e-06,
"loss": 0.0697,
"step": 706
},
{
"epoch": 0.05656,
"grad_norm": 0.13609936833381653,
"learning_rate": 4.8197299062687e-06,
"loss": 0.0266,
"step": 707
},
{
"epoch": 0.05664,
"grad_norm": 0.24375802278518677,
"learning_rate": 4.78990368159424e-06,
"loss": 0.0312,
"step": 708
},
{
"epoch": 0.05672,
"grad_norm": 0.974298894405365,
"learning_rate": 4.76014094029921e-06,
"loss": 0.0582,
"step": 709
},
{
"epoch": 0.0568,
"grad_norm": 0.2759767770767212,
"learning_rate": 4.7304420450332244e-06,
"loss": 0.033,
"step": 710
},
{
"epoch": 0.05688,
"grad_norm": 0.05576702207326889,
"learning_rate": 4.700807357667953e-06,
"loss": 0.0003,
"step": 711
},
{
"epoch": 0.05696,
"grad_norm": 0.1619577258825302,
"learning_rate": 4.671237239292699e-06,
"loss": 0.0316,
"step": 712
},
{
"epoch": 0.05704,
"grad_norm": 0.3270796239376068,
"learning_rate": 4.641732050210032e-06,
"loss": 0.0433,
"step": 713
},
{
"epoch": 0.05712,
"grad_norm": 1.177092432975769,
"learning_rate": 4.612292149931369e-06,
"loss": 0.063,
"step": 714
},
{
"epoch": 0.0572,
"grad_norm": 2.4695191383361816,
"learning_rate": 4.582917897172603e-06,
"loss": 0.0732,
"step": 715
},
{
"epoch": 0.05728,
"grad_norm": 0.027731506153941154,
"learning_rate": 4.5536096498497295e-06,
"loss": 0.0003,
"step": 716
},
{
"epoch": 0.05736,
"grad_norm": 0.2442253977060318,
"learning_rate": 4.524367765074499e-06,
"loss": 0.0291,
"step": 717
},
{
"epoch": 0.05744,
"grad_norm": 0.35619062185287476,
"learning_rate": 4.495192599150045e-06,
"loss": 0.0527,
"step": 718
},
{
"epoch": 0.05752,
"grad_norm": 1.317335844039917,
"learning_rate": 4.46608450756656e-06,
"loss": 0.0415,
"step": 719
},
{
"epoch": 0.0576,
"grad_norm": 0.7007994651794434,
"learning_rate": 4.437043844996952e-06,
"loss": 0.0394,
"step": 720
},
{
"epoch": 0.05768,
"grad_norm": 0.8067264556884766,
"learning_rate": 4.408070965292534e-06,
"loss": 0.0437,
"step": 721
},
{
"epoch": 0.05776,
"grad_norm": 0.46508219838142395,
"learning_rate": 4.379166221478697e-06,
"loss": 0.043,
"step": 722
},
{
"epoch": 0.05784,
"grad_norm": 0.5568006038665771,
"learning_rate": 4.350329965750622e-06,
"loss": 0.0471,
"step": 723
},
{
"epoch": 0.05792,
"grad_norm": 0.19141949713230133,
"learning_rate": 4.321562549468991e-06,
"loss": 0.0025,
"step": 724
},
{
"epoch": 0.058,
"grad_norm": 0.39139869809150696,
"learning_rate": 4.292864323155684e-06,
"loss": 0.0496,
"step": 725
},
{
"epoch": 0.05808,
"grad_norm": 0.3233785033226013,
"learning_rate": 4.264235636489542e-06,
"loss": 0.0352,
"step": 726
},
{
"epoch": 0.05816,
"grad_norm": 0.8876456022262573,
"learning_rate": 4.235676838302069e-06,
"loss": 0.0455,
"step": 727
},
{
"epoch": 0.05824,
"grad_norm": 0.30960965156555176,
"learning_rate": 4.207188276573214e-06,
"loss": 0.0432,
"step": 728
},
{
"epoch": 0.05832,
"grad_norm": 0.3074033260345459,
"learning_rate": 4.178770298427107e-06,
"loss": 0.0302,
"step": 729
},
{
"epoch": 0.0584,
"grad_norm": 0.35777154564857483,
"learning_rate": 4.150423250127846e-06,
"loss": 0.0348,
"step": 730
},
{
"epoch": 0.05848,
"grad_norm": 0.799038827419281,
"learning_rate": 4.12214747707527e-06,
"loss": 0.062,
"step": 731
},
{
"epoch": 0.05856,
"grad_norm": 1.9711394309997559,
"learning_rate": 4.093943323800746e-06,
"loss": 0.063,
"step": 732
},
{
"epoch": 0.05864,
"grad_norm": 0.3873506784439087,
"learning_rate": 4.065811133962987e-06,
"loss": 0.048,
"step": 733
},
{
"epoch": 0.05872,
"grad_norm": 0.4337482750415802,
"learning_rate": 4.037751250343841e-06,
"loss": 0.0531,
"step": 734
},
{
"epoch": 0.0588,
"grad_norm": 0.3944357633590698,
"learning_rate": 4.009764014844143e-06,
"loss": 0.0362,
"step": 735
},
{
"epoch": 0.05888,
"grad_norm": 0.7509347200393677,
"learning_rate": 3.981849768479516e-06,
"loss": 0.0479,
"step": 736
},
{
"epoch": 0.05896,
"grad_norm": 0.2694668173789978,
"learning_rate": 3.954008851376252e-06,
"loss": 0.0289,
"step": 737
},
{
"epoch": 0.05904,
"grad_norm": 0.02110792137682438,
"learning_rate": 3.9262416027671354e-06,
"loss": 0.0006,
"step": 738
},
{
"epoch": 0.05912,
"grad_norm": 0.24735328555107117,
"learning_rate": 3.898548360987325e-06,
"loss": 0.0277,
"step": 739
},
{
"epoch": 0.0592,
"grad_norm": 0.323829710483551,
"learning_rate": 3.8709294634702374e-06,
"loss": 0.0323,
"step": 740
},
{
"epoch": 0.05928,
"grad_norm": 0.8902094960212708,
"learning_rate": 3.8433852467434175e-06,
"loss": 0.019,
"step": 741
},
{
"epoch": 0.05936,
"grad_norm": 0.615180253982544,
"learning_rate": 3.81591604642446e-06,
"loss": 0.0611,
"step": 742
},
{
"epoch": 0.05944,
"grad_norm": 3.2819466590881348,
"learning_rate": 3.7885221972168974e-06,
"loss": 0.0388,
"step": 743
},
{
"epoch": 0.05952,
"grad_norm": 0.024337919428944588,
"learning_rate": 3.7612040329061405e-06,
"loss": 0.0003,
"step": 744
},
{
"epoch": 0.0596,
"grad_norm": 0.029666326940059662,
"learning_rate": 3.7339618863553983e-06,
"loss": 0.0003,
"step": 745
},
{
"epoch": 0.05968,
"grad_norm": 0.5997193455696106,
"learning_rate": 3.7067960895016277e-06,
"loss": 0.0423,
"step": 746
},
{
"epoch": 0.05976,
"grad_norm": 0.6478802561759949,
"learning_rate": 3.679706973351491e-06,
"loss": 0.0364,
"step": 747
},
{
"epoch": 0.05984,
"grad_norm": 0.3392534554004669,
"learning_rate": 3.6526948679773256e-06,
"loss": 0.0363,
"step": 748
},
{
"epoch": 0.05992,
"grad_norm": 0.5431677103042603,
"learning_rate": 3.625760102513103e-06,
"loss": 0.0491,
"step": 749
},
{
"epoch": 0.06,
"grad_norm": 0.2393576055765152,
"learning_rate": 3.598903005150444e-06,
"loss": 0.0095,
"step": 750
},
{
"epoch": 0.06008,
"grad_norm": 0.5906950235366821,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.0539,
"step": 751
},
{
"epoch": 0.06016,
"grad_norm": 0.40816617012023926,
"learning_rate": 3.545423122760493e-06,
"loss": 0.0438,
"step": 752
},
{
"epoch": 0.06024,
"grad_norm": 0.34278059005737305,
"learning_rate": 3.5188009893686916e-06,
"loss": 0.0541,
"step": 753
},
{
"epoch": 0.06032,
"grad_norm": 0.04328247159719467,
"learning_rate": 3.492257827341492e-06,
"loss": 0.001,
"step": 754
},
{
"epoch": 0.0604,
"grad_norm": 5.035195827484131,
"learning_rate": 3.4657939600989453e-06,
"loss": 0.0992,
"step": 755
},
{
"epoch": 0.06048,
"grad_norm": 0.6353648900985718,
"learning_rate": 3.4394097100949286e-06,
"loss": 0.0464,
"step": 756
},
{
"epoch": 0.06056,
"grad_norm": 1.5577665567398071,
"learning_rate": 3.4131053988131947e-06,
"loss": 0.0586,
"step": 757
},
{
"epoch": 0.06064,
"grad_norm": 0.024695999920368195,
"learning_rate": 3.3868813467634833e-06,
"loss": 0.0001,
"step": 758
},
{
"epoch": 0.06072,
"grad_norm": 0.018223987892270088,
"learning_rate": 3.360737873477584e-06,
"loss": 0.0004,
"step": 759
},
{
"epoch": 0.0608,
"grad_norm": 0.38315191864967346,
"learning_rate": 3.3346752975054763e-06,
"loss": 0.0474,
"step": 760
},
{
"epoch": 0.06088,
"grad_norm": 0.3016735017299652,
"learning_rate": 3.308693936411421e-06,
"loss": 0.0356,
"step": 761
},
{
"epoch": 0.06096,
"grad_norm": 1.048255205154419,
"learning_rate": 3.2827941067700996e-06,
"loss": 0.0562,
"step": 762
},
{
"epoch": 0.06104,
"grad_norm": 0.38259902596473694,
"learning_rate": 3.2569761241627694e-06,
"loss": 0.0331,
"step": 763
},
{
"epoch": 0.06112,
"grad_norm": 0.02355150692164898,
"learning_rate": 3.2312403031733943e-06,
"loss": 0.0004,
"step": 764
},
{
"epoch": 0.0612,
"grad_norm": 0.025226857513189316,
"learning_rate": 3.2055869573848374e-06,
"loss": 0.0002,
"step": 765
},
{
"epoch": 0.06128,
"grad_norm": 1.4983391761779785,
"learning_rate": 3.1800163993750166e-06,
"loss": 0.0533,
"step": 766
},
{
"epoch": 0.06136,
"grad_norm": 0.018019815906882286,
"learning_rate": 3.1545289407131128e-06,
"loss": 0.0003,
"step": 767
},
{
"epoch": 0.06144,
"grad_norm": 0.6588282585144043,
"learning_rate": 3.1291248919557717e-06,
"loss": 0.0612,
"step": 768
},
{
"epoch": 0.06152,
"grad_norm": 0.2595744729042053,
"learning_rate": 3.103804562643302e-06,
"loss": 0.0287,
"step": 769
},
{
"epoch": 0.0616,
"grad_norm": 0.8234947919845581,
"learning_rate": 3.0785682612959334e-06,
"loss": 0.0463,
"step": 770
},
{
"epoch": 0.06168,
"grad_norm": 0.3040834665298462,
"learning_rate": 3.0534162954100264e-06,
"loss": 0.035,
"step": 771
},
{
"epoch": 0.06176,
"grad_norm": 0.426418662071228,
"learning_rate": 3.028348971454356e-06,
"loss": 0.0494,
"step": 772
},
{
"epoch": 0.06184,
"grad_norm": 0.17616528272628784,
"learning_rate": 3.003366594866345e-06,
"loss": 0.0255,
"step": 773
},
{
"epoch": 0.06192,
"grad_norm": 0.18599098920822144,
"learning_rate": 2.978469470048376e-06,
"loss": 0.0221,
"step": 774
},
{
"epoch": 0.062,
"grad_norm": 0.5718640685081482,
"learning_rate": 2.953657900364053e-06,
"loss": 0.0625,
"step": 775
},
{
"epoch": 0.06208,
"grad_norm": 1.2851765155792236,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.0443,
"step": 776
},
{
"epoch": 0.06216,
"grad_norm": 0.14832432568073273,
"learning_rate": 2.9042926346347932e-06,
"loss": 0.0094,
"step": 777
},
{
"epoch": 0.06224,
"grad_norm": 0.6485694050788879,
"learning_rate": 2.8797395400900362e-06,
"loss": 0.0418,
"step": 778
},
{
"epoch": 0.06232,
"grad_norm": 0.5608707070350647,
"learning_rate": 2.855273203671969e-06,
"loss": 0.0262,
"step": 779
},
{
"epoch": 0.0624,
"grad_norm": 0.20021896064281464,
"learning_rate": 2.830893923495173e-06,
"loss": 0.0333,
"step": 780
},
{
"epoch": 0.06248,
"grad_norm": 0.14365260303020477,
"learning_rate": 2.8066019966134907e-06,
"loss": 0.0261,
"step": 781
},
{
"epoch": 0.06256,
"grad_norm": 0.020553817972540855,
"learning_rate": 2.7823977190163788e-06,
"loss": 0.0005,
"step": 782
},
{
"epoch": 0.06264,
"grad_norm": 0.029738636687397957,
"learning_rate": 2.7582813856253276e-06,
"loss": 0.0008,
"step": 783
},
{
"epoch": 0.06272,
"grad_norm": 0.43758726119995117,
"learning_rate": 2.7342532902902418e-06,
"loss": 0.0316,
"step": 784
},
{
"epoch": 0.0628,
"grad_norm": 0.2717728316783905,
"learning_rate": 2.7103137257858867e-06,
"loss": 0.0381,
"step": 785
},
{
"epoch": 0.06288,
"grad_norm": 0.32151567935943604,
"learning_rate": 2.6864629838082957e-06,
"loss": 0.0408,
"step": 786
},
{
"epoch": 0.06296,
"grad_norm": 0.22049684822559357,
"learning_rate": 2.6627013549712355e-06,
"loss": 0.0403,
"step": 787
},
{
"epoch": 0.06304,
"grad_norm": 1.853203535079956,
"learning_rate": 2.639029128802657e-06,
"loss": 0.1015,
"step": 788
},
{
"epoch": 0.06312,
"grad_norm": 2.5120890140533447,
"learning_rate": 2.615446593741161e-06,
"loss": 0.0314,
"step": 789
},
{
"epoch": 0.0632,
"grad_norm": 1.217475414276123,
"learning_rate": 2.5919540371325005e-06,
"loss": 0.0559,
"step": 790
},
{
"epoch": 0.06328,
"grad_norm": 0.25337734818458557,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0467,
"step": 791
},
{
"epoch": 0.06336,
"grad_norm": 0.2006894201040268,
"learning_rate": 2.5452400031713786e-06,
"loss": 0.0382,
"step": 792
},
{
"epoch": 0.06344,
"grad_norm": 0.3287632167339325,
"learning_rate": 2.522019095014683e-06,
"loss": 0.0511,
"step": 793
},
{
"epoch": 0.06352,
"grad_norm": 0.2655186057090759,
"learning_rate": 2.4988893036954045e-06,
"loss": 0.0372,
"step": 794
},
{
"epoch": 0.0636,
"grad_norm": 0.18994544446468353,
"learning_rate": 2.4758509110427576e-06,
"loss": 0.0277,
"step": 795
},
{
"epoch": 0.06368,
"grad_norm": 0.8951389193534851,
"learning_rate": 2.45290419777228e-06,
"loss": 0.0485,
"step": 796
},
{
"epoch": 0.06376,
"grad_norm": 0.012542092241346836,
"learning_rate": 2.4300494434824373e-06,
"loss": 0.0008,
"step": 797
},
{
"epoch": 0.06384,
"grad_norm": 0.4995518922805786,
"learning_rate": 2.407286926651192e-06,
"loss": 0.0419,
"step": 798
},
{
"epoch": 0.06392,
"grad_norm": 0.46922436356544495,
"learning_rate": 2.3846169246326345e-06,
"loss": 0.0487,
"step": 799
},
{
"epoch": 0.064,
"grad_norm": 0.4080369770526886,
"learning_rate": 2.362039713653581e-06,
"loss": 0.0438,
"step": 800
},
{
"epoch": 0.06408,
"grad_norm": 0.5322981476783752,
"learning_rate": 2.339555568810221e-06,
"loss": 0.043,
"step": 801
},
{
"epoch": 0.06416,
"grad_norm": 0.7284976840019226,
"learning_rate": 2.317164764064769e-06,
"loss": 0.0534,
"step": 802
},
{
"epoch": 0.06424,
"grad_norm": 1.3679851293563843,
"learning_rate": 2.2948675722421086e-06,
"loss": 0.0652,
"step": 803
},
{
"epoch": 0.06432,
"grad_norm": 0.17851702868938446,
"learning_rate": 2.27266426502649e-06,
"loss": 0.0271,
"step": 804
},
{
"epoch": 0.0644,
"grad_norm": 0.5474235415458679,
"learning_rate": 2.2505551129582047e-06,
"loss": 0.0548,
"step": 805
},
{
"epoch": 0.06448,
"grad_norm": 0.5281820893287659,
"learning_rate": 2.2285403854302912e-06,
"loss": 0.0394,
"step": 806
},
{
"epoch": 0.06456,
"grad_norm": 0.6997829079627991,
"learning_rate": 2.206620350685257e-06,
"loss": 0.0712,
"step": 807
},
{
"epoch": 0.06464,
"grad_norm": 0.3154629170894623,
"learning_rate": 2.1847952758118118e-06,
"loss": 0.0286,
"step": 808
},
{
"epoch": 0.06472,
"grad_norm": 0.3994235694408417,
"learning_rate": 2.163065426741603e-06,
"loss": 0.0433,
"step": 809
},
{
"epoch": 0.0648,
"grad_norm": 0.21205270290374756,
"learning_rate": 2.1414310682459805e-06,
"loss": 0.0419,
"step": 810
},
{
"epoch": 0.06488,
"grad_norm": 1.012418270111084,
"learning_rate": 2.119892463932781e-06,
"loss": 0.045,
"step": 811
},
{
"epoch": 0.06496,
"grad_norm": 1.3126273155212402,
"learning_rate": 2.098449876243096e-06,
"loss": 0.0512,
"step": 812
},
{
"epoch": 0.06504,
"grad_norm": 0.44224631786346436,
"learning_rate": 2.0771035664480944e-06,
"loss": 0.041,
"step": 813
},
{
"epoch": 0.06512,
"grad_norm": 0.2978363037109375,
"learning_rate": 2.0558537946458177e-06,
"loss": 0.0425,
"step": 814
},
{
"epoch": 0.0652,
"grad_norm": 0.2729344069957733,
"learning_rate": 2.0347008197580376e-06,
"loss": 0.0418,
"step": 815
},
{
"epoch": 0.06528,
"grad_norm": 0.022025760263204575,
"learning_rate": 2.013644899527074e-06,
"loss": 0.0007,
"step": 816
},
{
"epoch": 0.06536,
"grad_norm": 0.403136670589447,
"learning_rate": 1.9926862905126663e-06,
"loss": 0.0308,
"step": 817
},
{
"epoch": 0.06544,
"grad_norm": 0.06288518756628036,
"learning_rate": 1.9718252480888567e-06,
"loss": 0.0002,
"step": 818
},
{
"epoch": 0.06552,
"grad_norm": 1.0231736898422241,
"learning_rate": 1.95106202644086e-06,
"loss": 0.0244,
"step": 819
},
{
"epoch": 0.0656,
"grad_norm": 0.4465322494506836,
"learning_rate": 1.930396878561983e-06,
"loss": 0.0328,
"step": 820
},
{
"epoch": 0.06568,
"grad_norm": 0.3249387741088867,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.0337,
"step": 821
},
{
"epoch": 0.06576,
"grad_norm": 0.46426305174827576,
"learning_rate": 1.8893618101067357e-06,
"loss": 0.0323,
"step": 822
},
{
"epoch": 0.06584,
"grad_norm": 0.584151566028595,
"learning_rate": 1.8689923895297247e-06,
"loss": 0.0583,
"step": 823
},
{
"epoch": 0.06592,
"grad_norm": 0.02870512753725052,
"learning_rate": 1.848722042714457e-06,
"loss": 0.0003,
"step": 824
},
{
"epoch": 0.066,
"grad_norm": 0.1344195306301117,
"learning_rate": 1.8285510166487154e-06,
"loss": 0.0213,
"step": 825
},
{
"epoch": 0.06608,
"grad_norm": 0.21057887375354767,
"learning_rate": 1.808479557110081e-06,
"loss": 0.0285,
"step": 826
},
{
"epoch": 0.06616,
"grad_norm": 0.7634290456771851,
"learning_rate": 1.7885079086629598e-06,
"loss": 0.0397,
"step": 827
},
{
"epoch": 0.06624,
"grad_norm": 0.25600549578666687,
"learning_rate": 1.7686363146555807e-06,
"loss": 0.0273,
"step": 828
},
{
"epoch": 0.06632,
"grad_norm": 0.6897709369659424,
"learning_rate": 1.7488650172170496e-06,
"loss": 0.048,
"step": 829
},
{
"epoch": 0.0664,
"grad_norm": 1.406793236732483,
"learning_rate": 1.7291942572543806e-06,
"loss": 0.0417,
"step": 830
},
{
"epoch": 0.06648,
"grad_norm": 0.31128793954849243,
"learning_rate": 1.709624274449584e-06,
"loss": 0.0402,
"step": 831
},
{
"epoch": 0.06656,
"grad_norm": 0.2037745714187622,
"learning_rate": 1.6901553072567189e-06,
"loss": 0.0418,
"step": 832
},
{
"epoch": 0.06664,
"grad_norm": 0.9059445261955261,
"learning_rate": 1.6707875928990059e-06,
"loss": 0.0441,
"step": 833
},
{
"epoch": 0.06672,
"grad_norm": 0.4605213701725006,
"learning_rate": 1.651521367365936e-06,
"loss": 0.035,
"step": 834
},
{
"epoch": 0.0668,
"grad_norm": 0.47714921832084656,
"learning_rate": 1.6323568654103838e-06,
"loss": 0.0134,
"step": 835
},
{
"epoch": 0.06688,
"grad_norm": 0.5643584728240967,
"learning_rate": 1.6132943205457607e-06,
"loss": 0.0456,
"step": 836
},
{
"epoch": 0.06696,
"grad_norm": 0.24304045736789703,
"learning_rate": 1.5943339650431578e-06,
"loss": 0.0254,
"step": 837
},
{
"epoch": 0.06704,
"grad_norm": 0.43989163637161255,
"learning_rate": 1.5754760299285255e-06,
"loss": 0.0352,
"step": 838
},
{
"epoch": 0.06712,
"grad_norm": 0.21915112435817719,
"learning_rate": 1.5567207449798517e-06,
"loss": 0.0366,
"step": 839
},
{
"epoch": 0.0672,
"grad_norm": 0.6482210159301758,
"learning_rate": 1.538068338724361e-06,
"loss": 0.0425,
"step": 840
},
{
"epoch": 0.06728,
"grad_norm": 0.4032343626022339,
"learning_rate": 1.5195190384357405e-06,
"loss": 0.0628,
"step": 841
},
{
"epoch": 0.06736,
"grad_norm": 0.42773374915122986,
"learning_rate": 1.5010730701313626e-06,
"loss": 0.0599,
"step": 842
},
{
"epoch": 0.06744,
"grad_norm": 0.844366729259491,
"learning_rate": 1.4827306585695234e-06,
"loss": 0.0253,
"step": 843
},
{
"epoch": 0.06752,
"grad_norm": 1.200049638748169,
"learning_rate": 1.4644920272467245e-06,
"loss": 0.0564,
"step": 844
},
{
"epoch": 0.0676,
"grad_norm": 2.3227219581604004,
"learning_rate": 1.446357398394934e-06,
"loss": 0.0611,
"step": 845
},
{
"epoch": 0.06768,
"grad_norm": 0.49674665927886963,
"learning_rate": 1.4283269929788779e-06,
"loss": 0.0687,
"step": 846
},
{
"epoch": 0.06776,
"grad_norm": 0.7859125137329102,
"learning_rate": 1.4104010306933558e-06,
"loss": 0.0161,
"step": 847
},
{
"epoch": 0.06784,
"grad_norm": 3.1538383960723877,
"learning_rate": 1.3925797299605649e-06,
"loss": 0.0879,
"step": 848
},
{
"epoch": 0.06792,
"grad_norm": 1.3098032474517822,
"learning_rate": 1.3748633079274254e-06,
"loss": 0.0729,
"step": 849
},
{
"epoch": 0.068,
"grad_norm": 0.29121360182762146,
"learning_rate": 1.3572519804629537e-06,
"loss": 0.033,
"step": 850
},
{
"epoch": 0.06808,
"grad_norm": 0.5193185210227966,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0519,
"step": 851
},
{
"epoch": 0.06816,
"grad_norm": 0.29581573605537415,
"learning_rate": 1.322345466310717e-06,
"loss": 0.0483,
"step": 852
},
{
"epoch": 0.06824,
"grad_norm": 0.29322221875190735,
"learning_rate": 1.30505070494781e-06,
"loss": 0.0319,
"step": 853
},
{
"epoch": 0.06832,
"grad_norm": 0.5673242807388306,
"learning_rate": 1.2878618887981064e-06,
"loss": 0.0536,
"step": 854
},
{
"epoch": 0.0684,
"grad_norm": 1.3316998481750488,
"learning_rate": 1.2707792273019049e-06,
"loss": 0.0717,
"step": 855
},
{
"epoch": 0.06848,
"grad_norm": 0.3763214349746704,
"learning_rate": 1.2538029286060428e-06,
"loss": 0.0397,
"step": 856
},
{
"epoch": 0.06856,
"grad_norm": 3.0016326904296875,
"learning_rate": 1.2369331995613664e-06,
"loss": 0.0598,
"step": 857
},
{
"epoch": 0.06864,
"grad_norm": 0.44757550954818726,
"learning_rate": 1.2201702457201948e-06,
"loss": 0.0459,
"step": 858
},
{
"epoch": 0.06872,
"grad_norm": 0.2297152578830719,
"learning_rate": 1.2035142713338366e-06,
"loss": 0.0378,
"step": 859
},
{
"epoch": 0.0688,
"grad_norm": 1.1864118576049805,
"learning_rate": 1.1869654793500784e-06,
"loss": 0.0667,
"step": 860
},
{
"epoch": 0.06888,
"grad_norm": 0.1640380173921585,
"learning_rate": 1.1705240714107301e-06,
"loss": 0.0265,
"step": 861
},
{
"epoch": 0.06896,
"grad_norm": 0.4653007984161377,
"learning_rate": 1.1541902478491607e-06,
"loss": 0.0304,
"step": 862
},
{
"epoch": 0.06904,
"grad_norm": 0.25315096974372864,
"learning_rate": 1.1379642076878528e-06,
"loss": 0.0396,
"step": 863
},
{
"epoch": 0.06912,
"grad_norm": 0.45090773701667786,
"learning_rate": 1.1218461486359878e-06,
"loss": 0.0493,
"step": 864
},
{
"epoch": 0.0692,
"grad_norm": 0.5370141267776489,
"learning_rate": 1.1058362670870248e-06,
"loss": 0.032,
"step": 865
},
{
"epoch": 0.06928,
"grad_norm": 0.2245199978351593,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.028,
"step": 866
},
{
"epoch": 0.06936,
"grad_norm": 1.0205795764923096,
"learning_rate": 1.0741418154787443e-06,
"loss": 0.0418,
"step": 867
},
{
"epoch": 0.06944,
"grad_norm": 0.5019017457962036,
"learning_rate": 1.058457631606319e-06,
"loss": 0.0506,
"step": 868
},
{
"epoch": 0.06952,
"grad_norm": 0.5405016541481018,
"learning_rate": 1.042882397605871e-06,
"loss": 0.045,
"step": 869
},
{
"epoch": 0.0696,
"grad_norm": 1.0819233655929565,
"learning_rate": 1.0274163032567165e-06,
"loss": 0.0732,
"step": 870
},
{
"epoch": 0.06968,
"grad_norm": 0.30363285541534424,
"learning_rate": 1.012059537008332e-06,
"loss": 0.0037,
"step": 871
},
{
"epoch": 0.06976,
"grad_norm": 0.41418418288230896,
"learning_rate": 9.968122859780648e-07,
"loss": 0.0467,
"step": 872
},
{
"epoch": 0.06984,
"grad_norm": 0.6476432681083679,
"learning_rate": 9.816747359488632e-07,
"loss": 0.0489,
"step": 873
},
{
"epoch": 0.06992,
"grad_norm": 0.3859376907348633,
"learning_rate": 9.666470713669918e-07,
"loss": 0.0504,
"step": 874
},
{
"epoch": 0.07,
"grad_norm": 0.4879794120788574,
"learning_rate": 9.517294753398066e-07,
"loss": 0.0469,
"step": 875
},
{
"epoch": 0.07008,
"grad_norm": 1.1734009981155396,
"learning_rate": 9.369221296335007e-07,
"loss": 0.063,
"step": 876
},
{
"epoch": 0.07016,
"grad_norm": 0.9292811751365662,
"learning_rate": 9.222252146709143e-07,
"loss": 0.0611,
"step": 877
},
{
"epoch": 0.07024,
"grad_norm": 1.119032859802246,
"learning_rate": 9.076389095293148e-07,
"loss": 0.0539,
"step": 878
},
{
"epoch": 0.07032,
"grad_norm": 0.3730563521385193,
"learning_rate": 8.931633919382299e-07,
"loss": 0.0369,
"step": 879
},
{
"epoch": 0.0704,
"grad_norm": 0.17667949199676514,
"learning_rate": 8.787988382772705e-07,
"loss": 0.0363,
"step": 880
},
{
"epoch": 0.07048,
"grad_norm": 0.16174310445785522,
"learning_rate": 8.645454235739903e-07,
"loss": 0.0358,
"step": 881
},
{
"epoch": 0.07056,
"grad_norm": 0.2639504373073578,
"learning_rate": 8.504033215017527e-07,
"loss": 0.0373,
"step": 882
},
{
"epoch": 0.07064,
"grad_norm": 1.1223114728927612,
"learning_rate": 8.363727043776037e-07,
"loss": 0.0302,
"step": 883
},
{
"epoch": 0.07072,
"grad_norm": 1.8197335004806519,
"learning_rate": 8.224537431601886e-07,
"loss": 0.0733,
"step": 884
},
{
"epoch": 0.0708,
"grad_norm": 0.5685065388679504,
"learning_rate": 8.086466074476562e-07,
"loss": 0.0159,
"step": 885
},
{
"epoch": 0.07088,
"grad_norm": 0.7264477014541626,
"learning_rate": 7.949514654755963e-07,
"loss": 0.0442,
"step": 886
},
{
"epoch": 0.07096,
"grad_norm": 0.24182948470115662,
"learning_rate": 7.81368484114996e-07,
"loss": 0.0342,
"step": 887
},
{
"epoch": 0.07104,
"grad_norm": 0.2078809291124344,
"learning_rate": 7.678978288701911e-07,
"loss": 0.0292,
"step": 888
},
{
"epoch": 0.07112,
"grad_norm": 0.48908495903015137,
"learning_rate": 7.545396638768698e-07,
"loss": 0.0523,
"step": 889
},
{
"epoch": 0.0712,
"grad_norm": 0.4723975360393524,
"learning_rate": 7.412941519000527e-07,
"loss": 0.0357,
"step": 890
},
{
"epoch": 0.07128,
"grad_norm": 0.25341466069221497,
"learning_rate": 7.281614543321269e-07,
"loss": 0.0414,
"step": 891
},
{
"epoch": 0.07136,
"grad_norm": 0.9406892657279968,
"learning_rate": 7.151417311908648e-07,
"loss": 0.0486,
"step": 892
},
{
"epoch": 0.07144,
"grad_norm": 2.7003417015075684,
"learning_rate": 7.022351411174866e-07,
"loss": 0.052,
"step": 893
},
{
"epoch": 0.07152,
"grad_norm": 0.3869430720806122,
"learning_rate": 6.894418413747183e-07,
"loss": 0.0425,
"step": 894
},
{
"epoch": 0.0716,
"grad_norm": 0.935694694519043,
"learning_rate": 6.767619878448783e-07,
"loss": 0.0273,
"step": 895
},
{
"epoch": 0.07168,
"grad_norm": 0.3476157486438751,
"learning_rate": 6.641957350279838e-07,
"loss": 0.017,
"step": 896
},
{
"epoch": 0.07176,
"grad_norm": 0.6238088607788086,
"learning_rate": 6.517432360398556e-07,
"loss": 0.0644,
"step": 897
},
{
"epoch": 0.07184,
"grad_norm": 0.24337641894817352,
"learning_rate": 6.394046426102673e-07,
"loss": 0.0329,
"step": 898
},
{
"epoch": 0.07192,
"grad_norm": 0.22380399703979492,
"learning_rate": 6.271801050810856e-07,
"loss": 0.0438,
"step": 899
},
{
"epoch": 0.072,
"grad_norm": 0.665158748626709,
"learning_rate": 6.150697724044407e-07,
"loss": 0.0422,
"step": 900
},
{
"epoch": 0.07208,
"grad_norm": 0.5776143074035645,
"learning_rate": 6.030737921409169e-07,
"loss": 0.053,
"step": 901
},
{
"epoch": 0.07216,
"grad_norm": 0.2327408641576767,
"learning_rate": 5.911923104577455e-07,
"loss": 0.0345,
"step": 902
},
{
"epoch": 0.07224,
"grad_norm": 0.7090685963630676,
"learning_rate": 5.794254721270331e-07,
"loss": 0.0316,
"step": 903
},
{
"epoch": 0.07232,
"grad_norm": 0.35185936093330383,
"learning_rate": 5.677734205239904e-07,
"loss": 0.0371,
"step": 904
},
{
"epoch": 0.0724,
"grad_norm": 0.4443148970603943,
"learning_rate": 5.562362976251901e-07,
"loss": 0.0319,
"step": 905
},
{
"epoch": 0.07248,
"grad_norm": 0.35512015223503113,
"learning_rate": 5.448142440068316e-07,
"loss": 0.0416,
"step": 906
},
{
"epoch": 0.07256,
"grad_norm": 0.39638519287109375,
"learning_rate": 5.335073988430373e-07,
"loss": 0.0427,
"step": 907
},
{
"epoch": 0.07264,
"grad_norm": 0.5092903971672058,
"learning_rate": 5.223158999041444e-07,
"loss": 0.06,
"step": 908
},
{
"epoch": 0.07272,
"grad_norm": 0.3619992136955261,
"learning_rate": 5.112398835550348e-07,
"loss": 0.0477,
"step": 909
},
{
"epoch": 0.0728,
"grad_norm": 1.0755763053894043,
"learning_rate": 5.002794847534765e-07,
"loss": 0.0491,
"step": 910
},
{
"epoch": 0.07288,
"grad_norm": 0.3115963339805603,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0306,
"step": 911
},
{
"epoch": 0.07296,
"grad_norm": 0.3764609396457672,
"learning_rate": 4.787060725786141e-07,
"loss": 0.056,
"step": 912
},
{
"epoch": 0.07304,
"grad_norm": 0.7655561566352844,
"learning_rate": 4.6809332207053083e-07,
"loss": 0.0462,
"step": 913
},
{
"epoch": 0.07312,
"grad_norm": 1.5138109922409058,
"learning_rate": 4.575967148372318e-07,
"loss": 0.0534,
"step": 914
},
{
"epoch": 0.0732,
"grad_norm": 1.4939364194869995,
"learning_rate": 4.4721637877656377e-07,
"loss": 0.0525,
"step": 915
},
{
"epoch": 0.07328,
"grad_norm": 0.39590883255004883,
"learning_rate": 4.3695244036964567e-07,
"loss": 0.0423,
"step": 916
},
{
"epoch": 0.07336,
"grad_norm": 0.31334251165390015,
"learning_rate": 4.268050246793276e-07,
"loss": 0.0455,
"step": 917
},
{
"epoch": 0.07344,
"grad_norm": 0.26713865995407104,
"learning_rate": 4.167742553486676e-07,
"loss": 0.0422,
"step": 918
},
{
"epoch": 0.07352,
"grad_norm": 0.25376588106155396,
"learning_rate": 4.068602545994249e-07,
"loss": 0.0335,
"step": 919
},
{
"epoch": 0.0736,
"grad_norm": 0.7677134871482849,
"learning_rate": 3.9706314323056936e-07,
"loss": 0.0523,
"step": 920
},
{
"epoch": 0.07368,
"grad_norm": 2.359280586242676,
"learning_rate": 3.8738304061681107e-07,
"loss": 0.0743,
"step": 921
},
{
"epoch": 0.07376,
"grad_norm": 0.2769327461719513,
"learning_rate": 3.7782006470714614e-07,
"loss": 0.0427,
"step": 922
},
{
"epoch": 0.07384,
"grad_norm": 0.3320523798465729,
"learning_rate": 3.68374332023419e-07,
"loss": 0.0373,
"step": 923
},
{
"epoch": 0.07392,
"grad_norm": 1.6790692806243896,
"learning_rate": 3.590459576589e-07,
"loss": 0.0599,
"step": 924
},
{
"epoch": 0.074,
"grad_norm": 0.32433339953422546,
"learning_rate": 3.498350552768859e-07,
"loss": 0.0423,
"step": 925
},
{
"epoch": 0.07408,
"grad_norm": 0.702372670173645,
"learning_rate": 3.4074173710931804e-07,
"loss": 0.0452,
"step": 926
},
{
"epoch": 0.07416,
"grad_norm": 0.18786029517650604,
"learning_rate": 3.3176611395540625e-07,
"loss": 0.0413,
"step": 927
},
{
"epoch": 0.07424,
"grad_norm": 0.044344063848257065,
"learning_rate": 3.2290829518028867e-07,
"loss": 0.0001,
"step": 928
},
{
"epoch": 0.07432,
"grad_norm": 0.21276189386844635,
"learning_rate": 3.1416838871368925e-07,
"loss": 0.0274,
"step": 929
},
{
"epoch": 0.0744,
"grad_norm": 0.248125359416008,
"learning_rate": 3.0554650104861137e-07,
"loss": 0.043,
"step": 930
},
{
"epoch": 0.07448,
"grad_norm": 0.36459970474243164,
"learning_rate": 2.970427372400353e-07,
"loss": 0.0409,
"step": 931
},
{
"epoch": 0.07456,
"grad_norm": 0.47662386298179626,
"learning_rate": 2.8865720090364037e-07,
"loss": 0.0377,
"step": 932
},
{
"epoch": 0.07464,
"grad_norm": 0.879160463809967,
"learning_rate": 2.8038999421453827e-07,
"loss": 0.0251,
"step": 933
},
{
"epoch": 0.07472,
"grad_norm": 0.03137514740228653,
"learning_rate": 2.7224121790603517e-07,
"loss": 0.0004,
"step": 934
},
{
"epoch": 0.0748,
"grad_norm": 0.17797286808490753,
"learning_rate": 2.6421097126839714e-07,
"loss": 0.0274,
"step": 935
},
{
"epoch": 0.07488,
"grad_norm": 0.884548008441925,
"learning_rate": 2.5629935214764866e-07,
"loss": 0.0466,
"step": 936
},
{
"epoch": 0.07496,
"grad_norm": 0.9071290493011475,
"learning_rate": 2.4850645694436736e-07,
"loss": 0.0608,
"step": 937
},
{
"epoch": 0.07504,
"grad_norm": 0.5692517161369324,
"learning_rate": 2.4083238061252565e-07,
"loss": 0.0427,
"step": 938
},
{
"epoch": 0.07512,
"grad_norm": 0.3361085057258606,
"learning_rate": 2.332772166583208e-07,
"loss": 0.0431,
"step": 939
},
{
"epoch": 0.0752,
"grad_norm": 0.4243836998939514,
"learning_rate": 2.2584105713904126e-07,
"loss": 0.0345,
"step": 940
},
{
"epoch": 0.07528,
"grad_norm": 2.818598508834839,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.1325,
"step": 941
},
{
"epoch": 0.07536,
"grad_norm": 0.31802570819854736,
"learning_rate": 2.1132611238315004e-07,
"loss": 0.0352,
"step": 942
},
{
"epoch": 0.07544,
"grad_norm": 0.2141246348619461,
"learning_rate": 2.0424750400655947e-07,
"loss": 0.0297,
"step": 943
},
{
"epoch": 0.07552,
"grad_norm": 0.9887557625770569,
"learning_rate": 1.9728825378278248e-07,
"loss": 0.0595,
"step": 944
},
{
"epoch": 0.0756,
"grad_norm": 0.6057161092758179,
"learning_rate": 1.9044844650808468e-07,
"loss": 0.0437,
"step": 945
},
{
"epoch": 0.07568,
"grad_norm": 0.7023236751556396,
"learning_rate": 1.8372816552336025e-07,
"loss": 0.0585,
"step": 946
},
{
"epoch": 0.07576,
"grad_norm": 0.013023672625422478,
"learning_rate": 1.7712749271311392e-07,
"loss": 0.0009,
"step": 947
},
{
"epoch": 0.07584,
"grad_norm": 0.27686434984207153,
"learning_rate": 1.706465085044584e-07,
"loss": 0.0356,
"step": 948
},
{
"epoch": 0.07592,
"grad_norm": 0.9506880044937134,
"learning_rate": 1.6428529186614195e-07,
"loss": 0.0489,
"step": 949
},
{
"epoch": 0.076,
"grad_norm": 0.30680614709854126,
"learning_rate": 1.580439203075812e-07,
"loss": 0.0494,
"step": 950
},
{
"epoch": 0.07608,
"grad_norm": 0.17234249413013458,
"learning_rate": 1.519224698779198e-07,
"loss": 0.0436,
"step": 951
},
{
"epoch": 0.07616,
"grad_norm": 0.23512433469295502,
"learning_rate": 1.4592101516509916e-07,
"loss": 0.0345,
"step": 952
},
{
"epoch": 0.07624,
"grad_norm": 0.4689125716686249,
"learning_rate": 1.400396292949513e-07,
"loss": 0.0271,
"step": 953
},
{
"epoch": 0.07632,
"grad_norm": 1.3410956859588623,
"learning_rate": 1.3427838393030634e-07,
"loss": 0.0672,
"step": 954
},
{
"epoch": 0.0764,
"grad_norm": 0.2717195749282837,
"learning_rate": 1.2863734927012094e-07,
"loss": 0.0367,
"step": 955
},
{
"epoch": 0.07648,
"grad_norm": 0.44982394576072693,
"learning_rate": 1.231165940486234e-07,
"loss": 0.0346,
"step": 956
},
{
"epoch": 0.07656,
"grad_norm": 0.38558506965637207,
"learning_rate": 1.1771618553447217e-07,
"loss": 0.0397,
"step": 957
},
{
"epoch": 0.07664,
"grad_norm": 0.569456160068512,
"learning_rate": 1.1243618952994195e-07,
"loss": 0.031,
"step": 958
},
{
"epoch": 0.07672,
"grad_norm": 0.3015541732311249,
"learning_rate": 1.0727667037011668e-07,
"loss": 0.046,
"step": 959
},
{
"epoch": 0.0768,
"grad_norm": 0.36019936203956604,
"learning_rate": 1.0223769092211012e-07,
"loss": 0.0346,
"step": 960
},
{
"epoch": 0.07688,
"grad_norm": 0.6763150691986084,
"learning_rate": 9.731931258429638e-08,
"loss": 0.0254,
"step": 961
},
{
"epoch": 0.07696,
"grad_norm": 0.4306623637676239,
"learning_rate": 9.252159528556404e-08,
"loss": 0.0416,
"step": 962
},
{
"epoch": 0.07704,
"grad_norm": 0.5304408073425293,
"learning_rate": 8.784459748458318e-08,
"loss": 0.0516,
"step": 963
},
{
"epoch": 0.07712,
"grad_norm": 0.1952734887599945,
"learning_rate": 8.328837616909612e-08,
"loss": 0.0279,
"step": 964
},
{
"epoch": 0.0772,
"grad_norm": 0.6845470666885376,
"learning_rate": 7.885298685522235e-08,
"loss": 0.0668,
"step": 965
},
{
"epoch": 0.07728,
"grad_norm": 0.4335359036922455,
"learning_rate": 7.453848358678018e-08,
"loss": 0.0298,
"step": 966
},
{
"epoch": 0.07736,
"grad_norm": 0.42128944396972656,
"learning_rate": 7.034491893463059e-08,
"loss": 0.0467,
"step": 967
},
{
"epoch": 0.07744,
"grad_norm": 0.38508471846580505,
"learning_rate": 6.627234399603554e-08,
"loss": 0.0408,
"step": 968
},
{
"epoch": 0.07752,
"grad_norm": 0.3776645064353943,
"learning_rate": 6.232080839403631e-08,
"loss": 0.0507,
"step": 969
},
{
"epoch": 0.0776,
"grad_norm": 0.10946887731552124,
"learning_rate": 5.849036027684607e-08,
"loss": 0.0269,
"step": 970
},
{
"epoch": 0.07768,
"grad_norm": 0.018504632636904716,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0004,
"step": 971
},
{
"epoch": 0.07776,
"grad_norm": 0.4607369601726532,
"learning_rate": 5.119291171211793e-08,
"loss": 0.0457,
"step": 972
},
{
"epoch": 0.07784,
"grad_norm": 0.49736905097961426,
"learning_rate": 4.772600018168816e-08,
"loss": 0.0457,
"step": 973
},
{
"epoch": 0.07792,
"grad_norm": 0.025478772819042206,
"learning_rate": 4.438035396920004e-08,
"loss": 0.0003,
"step": 974
},
{
"epoch": 0.078,
"grad_norm": 1.4756689071655273,
"learning_rate": 4.115601384029666e-08,
"loss": 0.0741,
"step": 975
},
{
"epoch": 0.07808,
"grad_norm": 1.5168122053146362,
"learning_rate": 3.805301908254455e-08,
"loss": 0.0341,
"step": 976
},
{
"epoch": 0.07816,
"grad_norm": 0.015453828498721123,
"learning_rate": 3.50714075049563e-08,
"loss": 0.0008,
"step": 977
},
{
"epoch": 0.07824,
"grad_norm": 0.03477850928902626,
"learning_rate": 3.22112154375287e-08,
"loss": 0.0005,
"step": 978
},
{
"epoch": 0.07832,
"grad_norm": 0.48851799964904785,
"learning_rate": 2.947247773079753e-08,
"loss": 0.0131,
"step": 979
},
{
"epoch": 0.0784,
"grad_norm": 0.906639039516449,
"learning_rate": 2.6855227755419046e-08,
"loss": 0.0502,
"step": 980
},
{
"epoch": 0.07848,
"grad_norm": 0.023139460012316704,
"learning_rate": 2.4359497401758026e-08,
"loss": 0.0005,
"step": 981
},
{
"epoch": 0.07856,
"grad_norm": 0.03088991716504097,
"learning_rate": 2.1985317079500358e-08,
"loss": 0.0009,
"step": 982
},
{
"epoch": 0.07864,
"grad_norm": 0.4197922646999359,
"learning_rate": 1.973271571728441e-08,
"loss": 0.0499,
"step": 983
},
{
"epoch": 0.07872,
"grad_norm": 0.802412748336792,
"learning_rate": 1.7601720762346895e-08,
"loss": 0.0756,
"step": 984
},
{
"epoch": 0.0788,
"grad_norm": 0.46592894196510315,
"learning_rate": 1.5592358180189782e-08,
"loss": 0.065,
"step": 985
},
{
"epoch": 0.07888,
"grad_norm": 0.6646913886070251,
"learning_rate": 1.370465245426167e-08,
"loss": 0.0605,
"step": 986
},
{
"epoch": 0.07896,
"grad_norm": 0.15956738591194153,
"learning_rate": 1.1938626585660252e-08,
"loss": 0.0248,
"step": 987
},
{
"epoch": 0.07904,
"grad_norm": 0.2412548065185547,
"learning_rate": 1.0294302092853647e-08,
"loss": 0.035,
"step": 988
},
{
"epoch": 0.07912,
"grad_norm": 0.5365248322486877,
"learning_rate": 8.771699011416169e-09,
"loss": 0.0447,
"step": 989
},
{
"epoch": 0.0792,
"grad_norm": 1.0220736265182495,
"learning_rate": 7.370835893788508e-09,
"loss": 0.0434,
"step": 990
},
{
"epoch": 0.07928,
"grad_norm": 0.04382408410310745,
"learning_rate": 6.091729809042379e-09,
"loss": 0.0004,
"step": 991
},
{
"epoch": 0.07936,
"grad_norm": 1.0795965194702148,
"learning_rate": 4.9343963426840006e-09,
"loss": 0.0546,
"step": 992
},
{
"epoch": 0.07944,
"grad_norm": 0.02822449617087841,
"learning_rate": 3.898849596456477e-09,
"loss": 0.0003,
"step": 993
},
{
"epoch": 0.07952,
"grad_norm": 0.34624478220939636,
"learning_rate": 2.9851021881688314e-09,
"loss": 0.0441,
"step": 994
},
{
"epoch": 0.0796,
"grad_norm": 0.24021954834461212,
"learning_rate": 2.193165251545004e-09,
"loss": 0.0433,
"step": 995
},
{
"epoch": 0.07968,
"grad_norm": 0.5294962525367737,
"learning_rate": 1.5230484360873043e-09,
"loss": 0.0065,
"step": 996
},
{
"epoch": 0.07976,
"grad_norm": 1.0957331657409668,
"learning_rate": 9.74759906957612e-10,
"loss": 0.049,
"step": 997
},
{
"epoch": 0.07984,
"grad_norm": 0.22180429100990295,
"learning_rate": 5.483063448785686e-10,
"loss": 0.0255,
"step": 998
},
{
"epoch": 0.07992,
"grad_norm": 0.49117955565452576,
"learning_rate": 2.436929460525317e-10,
"loss": 0.0237,
"step": 999
},
{
"epoch": 0.08,
"grad_norm": 0.2544153928756714,
"learning_rate": 6.092342209607083e-11,
"loss": 0.0299,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1000,
"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": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}