e1_math_all_qwq / trainer_state.json
neginr's picture
End of training
068bfb2 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 9.121329296509744,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.2706,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 9.091319810463366,
"learning_rate": 6.451612903225807e-07,
"loss": 1.2582,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 9.135331196481932,
"learning_rate": 9.67741935483871e-07,
"loss": 1.2896,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 8.916646470147892,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.2755,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 8.488263164030808,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.2508,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 8.38388704815651,
"learning_rate": 1.935483870967742e-06,
"loss": 1.2584,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 6.860304824780966,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.1931,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 6.391848162152615,
"learning_rate": 2.580645161290323e-06,
"loss": 1.1816,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.8889196376270956,
"learning_rate": 2.903225806451613e-06,
"loss": 1.128,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 3.3919458600637036,
"learning_rate": 3.225806451612903e-06,
"loss": 1.0993,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 3.2334343420421585,
"learning_rate": 3.548387096774194e-06,
"loss": 1.1378,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 5.76978378368592,
"learning_rate": 3.870967741935484e-06,
"loss": 1.1161,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 6.082399984578812,
"learning_rate": 4.193548387096774e-06,
"loss": 1.1268,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 5.846847656332162,
"learning_rate": 4.516129032258065e-06,
"loss": 1.0947,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 5.895065437795678,
"learning_rate": 4.838709677419355e-06,
"loss": 1.0849,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 4.736228762058488,
"learning_rate": 5.161290322580646e-06,
"loss": 1.0193,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 4.482997670157339,
"learning_rate": 5.483870967741935e-06,
"loss": 0.9999,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 3.781630899926424,
"learning_rate": 5.806451612903226e-06,
"loss": 1.0165,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.628756839041789,
"learning_rate": 6.129032258064517e-06,
"loss": 0.9939,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.7658501517296634,
"learning_rate": 6.451612903225806e-06,
"loss": 0.9855,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.9497104660804716,
"learning_rate": 6.774193548387097e-06,
"loss": 0.9599,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.9937027129147626,
"learning_rate": 7.096774193548388e-06,
"loss": 0.9533,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.873406049913554,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.9248,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.5159082090482328,
"learning_rate": 7.741935483870968e-06,
"loss": 0.9127,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.360893640184476,
"learning_rate": 8.064516129032258e-06,
"loss": 0.9367,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.150850193484432,
"learning_rate": 8.387096774193549e-06,
"loss": 0.9171,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.0282217614129288,
"learning_rate": 8.70967741935484e-06,
"loss": 0.9143,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.8783927894611817,
"learning_rate": 9.03225806451613e-06,
"loss": 0.8706,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 1.0357909208741884,
"learning_rate": 9.35483870967742e-06,
"loss": 0.8756,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.8839724631117843,
"learning_rate": 9.67741935483871e-06,
"loss": 0.8944,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.8018392391703367,
"learning_rate": 1e-05,
"loss": 0.8884,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.9819998288383661,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.9039,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 1.043934971506977,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.8538,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.7486229782898178,
"learning_rate": 1.096774193548387e-05,
"loss": 0.8954,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.6462704913585466,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.8598,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.8701929511225565,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.8202,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.7286212857627473,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.8408,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6895180779108422,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.8474,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.590103429507888,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.8353,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.6938485731588618,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.8224,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.7356238898701509,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.8351,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.6897395139992688,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.8447,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.675549957371479,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.8281,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.705125192710144,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.8246,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.6759017317287515,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.8239,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.8411028689945631,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.8321,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.7811346752148822,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.8177,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.803634946287733,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.8274,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.743873737842097,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.8024,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.8063273957345632,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.8238,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.6106242273324267,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.7981,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.8876990632009988,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.7916,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.6101787978595211,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.8199,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.9544757529286129,
"learning_rate": 1.741935483870968e-05,
"loss": 0.812,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.793914153351299,
"learning_rate": 1.774193548387097e-05,
"loss": 0.8059,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.8558133984112029,
"learning_rate": 1.806451612903226e-05,
"loss": 0.8206,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 1.0363564468814344,
"learning_rate": 1.838709677419355e-05,
"loss": 0.8018,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.8425891932735051,
"learning_rate": 1.870967741935484e-05,
"loss": 0.7951,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 1.1899175260465815,
"learning_rate": 1.903225806451613e-05,
"loss": 0.7906,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.0448812772134173,
"learning_rate": 1.935483870967742e-05,
"loss": 0.8281,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.9603986761266647,
"learning_rate": 1.967741935483871e-05,
"loss": 0.8249,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 1.2302772948188017,
"learning_rate": 2e-05,
"loss": 0.7886,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.8593549639580391,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.7825,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.946141162856849,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.8096,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.8523513267178314,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.7958,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.9247888569623353,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.7742,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 1.149357869151732,
"learning_rate": 2.161290322580645e-05,
"loss": 0.7877,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.802823984381737,
"learning_rate": 2.193548387096774e-05,
"loss": 0.7821,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.9943040024539699,
"learning_rate": 2.225806451612903e-05,
"loss": 0.7925,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.7676683686443277,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.7951,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.2032439717283983,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.7975,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.8429334662941207,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.7938,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.996574786641722,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.7958,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.1550937692456882,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.761,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.9102510928298896,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.7797,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.7178703480996972,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.7834,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.8971996628211715,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.7947,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.0645159934129635,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.8123,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.0793870139126747,
"learning_rate": 2.548387096774194e-05,
"loss": 0.7946,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.0212181544818468,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.7829,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.313422271955504,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.776,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.2557571637923515,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.8055,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.1923247685606408,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.7918,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.190311449681825,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.7881,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.2128389087459592,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.7688,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.0821672748711315,
"learning_rate": 2.774193548387097e-05,
"loss": 0.7745,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.0148600899186964,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.775,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.082947560137704,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.7849,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.2560572930792835,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.7855,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.210533977958274,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.7835,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.1096244485588531,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.7516,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.362162037427774,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.7655,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.1987086677385113,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.7758,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.1879289312969237,
"learning_rate": 3.032258064516129e-05,
"loss": 0.8187,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.4088333305756107,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.7955,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.9079648373654176,
"learning_rate": 3.096774193548387e-05,
"loss": 0.7723,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.3649822979495143,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.7653,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.341688187887982,
"learning_rate": 3.161290322580645e-05,
"loss": 0.7617,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.059275707045646,
"learning_rate": 3.193548387096774e-05,
"loss": 0.7745,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.8668461869684878,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.7985,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.2618306847834149,
"learning_rate": 3.258064516129033e-05,
"loss": 0.7725,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.7494194917068926,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.7721,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.7128754679849454,
"learning_rate": 3.322580645161291e-05,
"loss": 0.7996,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.1693352213594277,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.7788,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.4731573984836415,
"learning_rate": 3.387096774193549e-05,
"loss": 0.7989,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.1165684953566748,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.7869,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.1186868452248007,
"learning_rate": 3.451612903225807e-05,
"loss": 0.787,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.2715269641458047,
"learning_rate": 3.483870967741936e-05,
"loss": 0.7369,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.0185823021400118,
"learning_rate": 3.516129032258065e-05,
"loss": 0.7736,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.1976918670059749,
"learning_rate": 3.548387096774194e-05,
"loss": 0.7665,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.3244810847509667,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.7926,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.1649758533936228,
"learning_rate": 3.612903225806452e-05,
"loss": 0.7883,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.8954477667180475,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.7552,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.2696244182565921,
"learning_rate": 3.67741935483871e-05,
"loss": 0.7483,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.0973184360563406,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.7614,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.7697269017005476,
"learning_rate": 3.741935483870968e-05,
"loss": 0.7503,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 1.0132560803966664,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.79,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.1833583860077874,
"learning_rate": 3.806451612903226e-05,
"loss": 0.7859,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.956445658510701,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.772,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.088815345553717,
"learning_rate": 3.870967741935484e-05,
"loss": 0.76,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.4916946630279013,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.7697,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.121458854005987,
"learning_rate": 3.935483870967742e-05,
"loss": 0.7672,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.0565482527466548,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.7659,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.6746169764955867,
"learning_rate": 4e-05,
"loss": 0.7787,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.0383355901792108,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.7658,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.754469374696345,
"learning_rate": 3.999968016170525e-05,
"loss": 0.7992,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.2203914118301802,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.7848,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.3066183623453955,
"learning_rate": 3.999872065705065e-05,
"loss": 0.76,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.0854328307631824,
"learning_rate": 3.999800103862945e-05,
"loss": 0.7453,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.248537054306766,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.759,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.9720389937172301,
"learning_rate": 3.999608209836944e-05,
"loss": 0.7524,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.9853333335353657,
"learning_rate": 3.999488279187443e-05,
"loss": 0.7537,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.2852430652972304,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.7377,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.0177530493223197,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.7883,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.1137489480632317,
"learning_rate": 3.999032564583976e-05,
"loss": 0.7586,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.4139803783883036,
"learning_rate": 3.998848689546591e-05,
"loss": 0.7637,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.9542409108654848,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.7608,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.2203309704530232,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.7619,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 1.1557529748169442,
"learning_rate": 3.998201174509306e-05,
"loss": 0.7824,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.9793869374242515,
"learning_rate": 3.997953378607961e-05,
"loss": 0.7983,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 1.0903453530462037,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.7884,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.9291956122733006,
"learning_rate": 3.997409862168454e-05,
"loss": 0.7416,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 1.016308927912477,
"learning_rate": 3.997114145976237e-05,
"loss": 0.7632,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.8015803231040982,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.7612,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.9184525752610007,
"learning_rate": 3.996474809469545e-05,
"loss": 0.7686,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.8718348761191019,
"learning_rate": 3.996131194267189e-05,
"loss": 0.7482,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.7150771244383564,
"learning_rate": 3.995771618052995e-05,
"loss": 0.7461,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.8082217052696498,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.7815,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.8430472233827607,
"learning_rate": 3.995004594217343e-05,
"loss": 0.7586,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.9719152359407901,
"learning_rate": 3.994597152728987e-05,
"loss": 0.7777,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.7664287768516441,
"learning_rate": 3.994173762494947e-05,
"loss": 0.7476,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.0261495325539534,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.7728,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.6671840467815392,
"learning_rate": 3.993279149458988e-05,
"loss": 0.7762,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.7009273270197129,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.7484,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.7832918990279965,
"learning_rate": 3.992320783722616e-05,
"loss": 0.7383,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.6736108510234309,
"learning_rate": 3.991817703090965e-05,
"loss": 0.7564,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.6727880124531556,
"learning_rate": 3.991298695938038e-05,
"loss": 0.7854,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.8407287505064291,
"learning_rate": 3.9907637664138e-05,
"loss": 0.7477,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 1.006090889814394,
"learning_rate": 3.990212918795534e-05,
"loss": 0.7648,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.8963092571066855,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.7325,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.9404330221907863,
"learning_rate": 3.989063487022416e-05,
"loss": 0.768,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.6625104652903973,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.7532,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.8559194797251429,
"learning_rate": 3.987850437381915e-05,
"loss": 0.7669,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.9806968012034305,
"learning_rate": 3.987220067906313e-05,
"loss": 0.7468,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.6431915172030198,
"learning_rate": 3.986573808672002e-05,
"loss": 0.7504,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.9606860062504982,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.751,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 1.2856459991331608,
"learning_rate": 3.985233641724152e-05,
"loss": 0.7563,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.7680530723008928,
"learning_rate": 3.984539744726554e-05,
"loss": 0.7515,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 1.3759242517249415,
"learning_rate": 3.983829979402038e-05,
"loss": 0.7675,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.9473792799882074,
"learning_rate": 3.983104351425871e-05,
"loss": 0.7551,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 1.18835152810261,
"learning_rate": 3.982362866600154e-05,
"loss": 0.7463,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.9045934415596437,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.7506,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 1.0009123550032815,
"learning_rate": 3.980832350242388e-05,
"loss": 0.741,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 1.004205725621572,
"learning_rate": 3.980043330948306e-05,
"loss": 0.7698,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.8936021549059571,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.7765,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.7035523733676269,
"learning_rate": 3.978417801674579e-05,
"loss": 0.7649,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.731617281607026,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.7857,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.6372913067253148,
"learning_rate": 3.976728995023254e-05,
"loss": 0.7171,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.8259807652854149,
"learning_rate": 3.975860879481514e-05,
"loss": 0.7657,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.6601476190272947,
"learning_rate": 3.974976965008833e-05,
"loss": 0.7583,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.727217087747953,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.7581,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.6953933042294929,
"learning_rate": 3.973161767667946e-05,
"loss": 0.7489,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.7659061897635185,
"learning_rate": 3.972230499314009e-05,
"loss": 0.7541,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.7978549692754017,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.7447,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.6987669365924369,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.7497,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.7302875706584051,
"learning_rate": 3.969342105253098e-05,
"loss": 0.759,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.7611469435459921,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.7677,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.6864666017617235,
"learning_rate": 3.967337762346568e-05,
"loss": 0.7885,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.7657302852845205,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.7679,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.7502466977783803,
"learning_rate": 3.965270496444528e-05,
"loss": 0.7147,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.7603149659191784,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.7681,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.6593532080051538,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.7477,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.8510550675506099,
"learning_rate": 3.962051762160151e-05,
"loss": 0.7633,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.691383718396142,
"learning_rate": 3.960947462140638e-05,
"loss": 0.7364,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.6510885950451669,
"learning_rate": 3.959827482437472e-05,
"loss": 0.7629,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6908707750236313,
"learning_rate": 3.958691832005981e-05,
"loss": 0.7396,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.7081685349470226,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.7513,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.6021129030878707,
"learning_rate": 3.956373555405774e-05,
"loss": 0.7471,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.9708921725850689,
"learning_rate": 3.955190947773937e-05,
"loss": 0.7672,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.8957213690324696,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.7641,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.93281171462015,
"learning_rate": 3.952778841127214e-05,
"loss": 0.7451,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.63794562132054,
"learning_rate": 3.95154936139947e-05,
"loss": 0.7408,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.7665698346093501,
"learning_rate": 3.950304277135034e-05,
"loss": 0.7392,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.7258131438391378,
"learning_rate": 3.94904359828957e-05,
"loss": 0.7297,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.7415638674156643,
"learning_rate": 3.94776733494343e-05,
"loss": 0.7613,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.8087091741904863,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.7502,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.6603603993711413,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.7412,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.8822260722604313,
"learning_rate": 3.943845140573213e-05,
"loss": 0.7491,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.7956329868015921,
"learning_rate": 3.942506642518952e-05,
"loss": 0.7746,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.5895681889391508,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.7282,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.8449665601244362,
"learning_rate": 3.939783060543167e-05,
"loss": 0.7776,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.784255825488041,
"learning_rate": 3.938397998399333e-05,
"loss": 0.7554,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.7108305036263703,
"learning_rate": 3.936997436876756e-05,
"loss": 0.7482,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.6908837331548332,
"learning_rate": 3.935581387174289e-05,
"loss": 0.755,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.8607774778309991,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.7638,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.6172923038395287,
"learning_rate": 3.932702868644224e-05,
"loss": 0.7434,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.6436710797727847,
"learning_rate": 3.931240422833181e-05,
"loss": 0.7406,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.6221948266356285,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.7266,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.5816477140130917,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.7308,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6922429816759977,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.718,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.6554774039827584,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.7349,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.8755381151245861,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.7303,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.7211683309379313,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.7631,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.7012342094546198,
"learning_rate": 3.92057159243286e-05,
"loss": 0.7471,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.6251607331692587,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.7538,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.7979607705144571,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.7647,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.633827828956145,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.7509,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.7217042094435436,
"learning_rate": 3.914136997982874e-05,
"loss": 0.7377,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.6648622062994693,
"learning_rate": 3.912490053436079e-05,
"loss": 0.7328,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.6439702394799871,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.7418,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.7829609985841033,
"learning_rate": 3.909150300975178e-05,
"loss": 0.7351,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.7755096689988501,
"learning_rate": 3.907457519765642e-05,
"loss": 0.7321,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.6553524509920873,
"learning_rate": 3.905749486576607e-05,
"loss": 0.7827,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.7955216206743478,
"learning_rate": 3.90402621506546e-05,
"loss": 0.7124,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.5620415222511096,
"learning_rate": 3.902287719011435e-05,
"loss": 0.7372,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.750702545158002,
"learning_rate": 3.900534012315499e-05,
"loss": 0.7338,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.6453934819765296,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.7361,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.7407773842619034,
"learning_rate": 3.896981023209777e-05,
"loss": 0.7507,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.6043990238608433,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.7321,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.6039718681215387,
"learning_rate": 3.893367361386491e-05,
"loss": 0.7321,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.5561285071974676,
"learning_rate": 3.891537814248417e-05,
"loss": 0.7504,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.6260147304650826,
"learning_rate": 3.889693142424384e-05,
"loss": 0.741,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.5151915953775488,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.7144,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.5864366380445825,
"learning_rate": 3.885958483839049e-05,
"loss": 0.7373,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.5415953173852421,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.746,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.591366744373595,
"learning_rate": 3.882163505079171e-05,
"loss": 0.7608,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.5557211620325473,
"learning_rate": 3.880243433489132e-05,
"loss": 0.7388,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.5932316342949188,
"learning_rate": 3.878308327522702e-05,
"loss": 0.6796,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.7606101918438111,
"learning_rate": 3.876358202652936e-05,
"loss": 0.6918,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.7714278574986159,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.6731,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.8212429768673442,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.6809,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.618836489262366,
"learning_rate": 3.870417871154801e-05,
"loss": 0.6727,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.7740836785057976,
"learning_rate": 3.868407827802192e-05,
"loss": 0.6655,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.816422125736489,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.6583,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.6419984810025379,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.6573,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.760909993218476,
"learning_rate": 3.862288124195319e-05,
"loss": 0.6604,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.7229464870506127,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.6856,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.5939254359110295,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.6765,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.6150086159762632,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.6821,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.5653991203589832,
"learning_rate": 3.853920126717703e-05,
"loss": 0.6646,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.5807068775284439,
"learning_rate": 3.851791025492574e-05,
"loss": 0.6638,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.6126700445645195,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.65,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.7950891404812495,
"learning_rate": 3.847488419569866e-05,
"loss": 0.6765,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.6140361587278209,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.6643,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.5831040225364098,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.6586,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.7042014690648415,
"learning_rate": 3.840923760917199e-05,
"loss": 0.6903,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.5717518336820003,
"learning_rate": 3.838706077964473e-05,
"loss": 0.6608,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.6275331516838298,
"learning_rate": 3.836473692766944e-05,
"loss": 0.6976,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.49067421458553423,
"learning_rate": 3.834226623174705e-05,
"loss": 0.6814,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.5596904789637558,
"learning_rate": 3.831964887155264e-05,
"loss": 0.6484,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.5640386254888526,
"learning_rate": 3.829688502793403e-05,
"loss": 0.6791,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.5637086373530383,
"learning_rate": 3.827397488291029e-05,
"loss": 0.6672,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.5339299489582772,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.6585,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.6009918746172792,
"learning_rate": 3.822771642257145e-05,
"loss": 0.6693,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.4862483740469908,
"learning_rate": 3.820436847713776e-05,
"loss": 0.679,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.5139905935274418,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.6611,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.5224308376677845,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.6828,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.5057327251040419,
"learning_rate": 3.813345202354145e-05,
"loss": 0.6664,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.5298382613813024,
"learning_rate": 3.810952296329563e-05,
"loss": 0.6681,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.5932751138790933,
"learning_rate": 3.808544909978676e-05,
"loss": 0.6733,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.545199528050084,
"learning_rate": 3.806123062550882e-05,
"loss": 0.6735,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.6470799291780789,
"learning_rate": 3.803686773411208e-05,
"loss": 0.6528,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.6227545339824304,
"learning_rate": 3.801236062040158e-05,
"loss": 0.6937,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.6032178369380694,
"learning_rate": 3.798770948033553e-05,
"loss": 0.6642,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.7253791019614426,
"learning_rate": 3.796291451102381e-05,
"loss": 0.6854,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.6100037867606768,
"learning_rate": 3.793797591072631e-05,
"loss": 0.6635,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.6400696053347493,
"learning_rate": 3.791289387885144e-05,
"loss": 0.6855,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.5798976775853734,
"learning_rate": 3.788766861595443e-05,
"loss": 0.6732,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.6605181198818364,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.6695,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.6085600493063897,
"learning_rate": 3.783678920503983e-05,
"loss": 0.6633,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.6286948067802067,
"learning_rate": 3.781113546385265e-05,
"loss": 0.6763,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.6600863381076506,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.6844,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.6057155507945885,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.6649,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.5583984600958903,
"learning_rate": 3.77333205623025e-05,
"loss": 0.6788,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.6428761468228736,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.6566,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.5432410314281388,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.6842,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.6421954275759026,
"learning_rate": 3.765422951112704e-05,
"loss": 0.663,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.5557109056070239,
"learning_rate": 3.762758321970216e-05,
"loss": 0.6799,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.6209247352922355,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.6801,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.6293032440151622,
"learning_rate": 3.757386800198538e-05,
"loss": 0.6827,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.6724512207664797,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.7073,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.5332693656258675,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.6739,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.6730592634763586,
"learning_rate": 3.749224181796305e-05,
"loss": 0.7064,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5897166206948851,
"learning_rate": 3.746475306375545e-05,
"loss": 0.6748,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.67807630061184,
"learning_rate": 3.743712466184774e-05,
"loss": 0.6768,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.5254961270177008,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.6821,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.6561536629160707,
"learning_rate": 3.73814497997107e-05,
"loss": 0.6772,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.5618943870520086,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.6847,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.6246300809726967,
"learning_rate": 3.732521901224724e-05,
"loss": 0.691,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.5743440267378209,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.6866,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.5653866423578602,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.672,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.579070060405949,
"learning_rate": 3.723983441007882e-05,
"loss": 0.6775,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.7039868314632145,
"learning_rate": 3.721109687296781e-05,
"loss": 0.6756,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.5108820469359523,
"learning_rate": 3.718222171638484e-05,
"loss": 0.6645,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.6050267011990201,
"learning_rate": 3.715320917121488e-05,
"loss": 0.6731,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.5076895551467205,
"learning_rate": 3.712405946944149e-05,
"loss": 0.659,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.5205577733601737,
"learning_rate": 3.709477284414488e-05,
"loss": 0.6678,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.5383477841447166,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.6678,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.5389617611475348,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.6736,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.652941338322821,
"learning_rate": 3.70060937743295e-05,
"loss": 0.6638,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.5091211748195853,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.676,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.8043779498333848,
"learning_rate": 3.694629409915555e-05,
"loss": 0.6497,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.5101189781388719,
"learning_rate": 3.691619088858401e-05,
"loss": 0.6825,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.6206533947919837,
"learning_rate": 3.68859524166009e-05,
"loss": 0.6501,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.5837773305695467,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.6914,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.6787569573223757,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.6634,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.5341135305081771,
"learning_rate": 3.67944278554384e-05,
"loss": 0.6965,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.5562465500124067,
"learning_rate": 3.676365076645557e-05,
"loss": 0.6849,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.45752770000483817,
"learning_rate": 3.673273963576791e-05,
"loss": 0.6591,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.5276216797722513,
"learning_rate": 3.670169471054001e-05,
"loss": 0.6802,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.47128177810447774,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.6829,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.5297555852990001,
"learning_rate": 3.663920447046888e-05,
"loss": 0.678,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.5303511196810328,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.668,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.5394514680925842,
"learning_rate": 3.657618204491937e-05,
"loss": 0.701,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.48168093052444066,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.6472,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.6537932167947016,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.6775,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.512968679891127,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.6826,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.7529682557677772,
"learning_rate": 3.644854871713611e-05,
"loss": 0.6767,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.5974497403899812,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.6858,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.5286110012643073,
"learning_rate": 3.638394189710493e-05,
"loss": 0.6462,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.47255124574000656,
"learning_rate": 3.635144184933412e-05,
"loss": 0.6672,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5283445615738724,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.6771,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.5456168896891174,
"learning_rate": 3.628604977762752e-05,
"loss": 0.6802,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.5911589547942566,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.6587,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.6396974986319158,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.6938,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.5151428083565206,
"learning_rate": 3.618698565901727e-05,
"loss": 0.6818,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.5762157561549147,
"learning_rate": 3.615370507164654e-05,
"loss": 0.6715,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.5226898049785967,
"learning_rate": 3.61202953196805e-05,
"loss": 0.6725,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.4865334854402519,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.6693,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.5791838422319145,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.6885,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5061301349472137,
"learning_rate": 3.601929375279662e-05,
"loss": 0.6693,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.7130126618336983,
"learning_rate": 3.598537002418012e-05,
"loss": 0.6707,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.6783553289180062,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.6841,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.6107282371680716,
"learning_rate": 3.591713938344416e-05,
"loss": 0.6778,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.8544472066549508,
"learning_rate": 3.588283301689507e-05,
"loss": 0.6906,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.624684071266017,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.6834,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.6322129250685512,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.6674,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.5968739771854935,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.6914,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.5953146769042917,
"learning_rate": 3.574434032194994e-05,
"loss": 0.6597,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.5576257631450988,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.672,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.6088977877291853,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.6682,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.4514727212860745,
"learning_rate": 3.563914797637216e-05,
"loss": 0.6739,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.6042724738131604,
"learning_rate": 3.560383338592649e-05,
"loss": 0.6774,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.558515124105638,
"learning_rate": 3.556839402764487e-05,
"loss": 0.6811,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.4850028669316893,
"learning_rate": 3.553283018489946e-05,
"loss": 0.6793,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.5475320749196037,
"learning_rate": 3.54971421420578e-05,
"loss": 0.6741,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.5049667863547536,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.6866,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.47888111635930736,
"learning_rate": 3.54253945985191e-05,
"loss": 0.6743,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.4729435005233576,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.6707,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.484923396335231,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.6556,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.46932513331192555,
"learning_rate": 3.531684894865866e-05,
"loss": 0.671,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.5285497172839391,
"learning_rate": 3.528042173241128e-05,
"loss": 0.6846,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.5708864102531406,
"learning_rate": 3.524387233431892e-05,
"loss": 0.6722,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.46082288799816123,
"learning_rate": 3.520720104662958e-05,
"loss": 0.6761,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.5505831773847523,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.6752,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.5113048357675176,
"learning_rate": 3.513349397632281e-05,
"loss": 0.6589,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.5576467995177203,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.6647,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.5311525471693501,
"learning_rate": 3.505930287892533e-05,
"loss": 0.6743,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.5756032518488177,
"learning_rate": 3.502202656100099e-05,
"loss": 0.6468,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.535453728960118,
"learning_rate": 3.498463012735256e-05,
"loss": 0.6956,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.4904466186861154,
"learning_rate": 3.494711387700094e-05,
"loss": 0.6923,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.49473474149142194,
"learning_rate": 3.490947810992505e-05,
"loss": 0.6671,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.4899440807438458,
"learning_rate": 3.48717231270595e-05,
"loss": 0.6674,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.5572335805360075,
"learning_rate": 3.483384923029211e-05,
"loss": 0.7007,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.4613782345493894,
"learning_rate": 3.479585672246156e-05,
"loss": 0.6723,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.4832978443822716,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.68,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.4688160265335252,
"learning_rate": 3.471951708970527e-05,
"loss": 0.6855,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.5589600051928013,
"learning_rate": 3.468117057518921e-05,
"loss": 0.6692,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.483413356329762,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.6656,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.633829475121958,
"learning_rate": 3.460412568296731e-05,
"loss": 0.664,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.5580284231491257,
"learning_rate": 3.456542792131039e-05,
"loss": 0.6901,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.4447818519866122,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.6725,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.5595881601325845,
"learning_rate": 3.448768331403348e-05,
"loss": 0.6647,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.45824836547469006,
"learning_rate": 3.44486370900573e-05,
"loss": 0.6887,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.4958582976279086,
"learning_rate": 3.440947533516396e-05,
"loss": 0.6499,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.5113429126313062,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.6965,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.432093753041472,
"learning_rate": 3.43308064860925e-05,
"loss": 0.6718,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.5891591232149066,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.6579,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.47788022785648204,
"learning_rate": 3.425167928295015e-05,
"loss": 0.6617,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.4942368593802549,
"learning_rate": 3.421194458890411e-05,
"loss": 0.6816,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.4862842402251212,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.6613,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.43560795400446867,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.6655,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.5987341850894032,
"learning_rate": 3.409205995219561e-05,
"loss": 0.6647,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.49059147549556675,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.6922,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.5077786204890812,
"learning_rate": 3.40115729298209e-05,
"loss": 0.6771,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.5914030760276526,
"learning_rate": 3.397116120327025e-05,
"loss": 0.6752,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.500240051875466,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.6737,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.5577639190534911,
"learning_rate": 3.389000293509524e-05,
"loss": 0.6791,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.5138295652435717,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.6802,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.5189651042105784,
"learning_rate": 3.380840041143494e-05,
"loss": 0.6604,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.5080595664152389,
"learning_rate": 3.376743336885792e-05,
"loss": 0.6718,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.525525096893852,
"learning_rate": 3.372635624225055e-05,
"loss": 0.6848,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.4986007729106607,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.6898,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.54904775185534,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.6541,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.4361052864317557,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.6615,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.6359530997618446,
"learning_rate": 3.356095347769799e-05,
"loss": 0.6795,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.5204428555421495,
"learning_rate": 3.351933087522552e-05,
"loss": 0.6635,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.590003239777369,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.6574,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.7302061624958971,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.6867,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.5870702288545724,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.6783,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.6162426156217341,
"learning_rate": 3.335176280432308e-05,
"loss": 0.6822,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.6055467581514804,
"learning_rate": 3.330960304618295e-05,
"loss": 0.7019,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.5066043242726498,
"learning_rate": 3.326733686481153e-05,
"loss": 0.6836,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.5682249580789255,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.6716,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.4859169273507345,
"learning_rate": 3.318248658506012e-05,
"loss": 0.662,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.6644478278982128,
"learning_rate": 3.313990316514069e-05,
"loss": 0.6886,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.4618137458889507,
"learning_rate": 3.309721467890571e-05,
"loss": 0.6676,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.5717899163870815,
"learning_rate": 3.305442146769115e-05,
"loss": 0.6678,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.44497964926218503,
"learning_rate": 3.30115238736704e-05,
"loss": 0.6399,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.5992253937309523,
"learning_rate": 3.296852223985148e-05,
"loss": 0.6803,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.5183216459355396,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.6835,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.5000793937867666,
"learning_rate": 3.288220822900796e-05,
"loss": 0.6556,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.5791869168776654,
"learning_rate": 3.283889654214788e-05,
"loss": 0.6729,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.5685367906426194,
"learning_rate": 3.279548219581317e-05,
"loss": 0.7066,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.5054174675877985,
"learning_rate": 3.275196553714379e-05,
"loss": 0.6691,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.4503000973648763,
"learning_rate": 3.270834691409778e-05,
"loss": 0.6683,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.5174611912214043,
"learning_rate": 3.266462667544849e-05,
"loss": 0.6431,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.502454119624543,
"learning_rate": 3.262080517078178e-05,
"loss": 0.6853,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.5635968744160529,
"learning_rate": 3.257688275049323e-05,
"loss": 0.677,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.4785024710188593,
"learning_rate": 3.253285976578535e-05,
"loss": 0.6642,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.5185497488883148,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.6553,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.44867932620790557,
"learning_rate": 3.244451351193935e-05,
"loss": 0.6676,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.5074569302282075,
"learning_rate": 3.240019094921551e-05,
"loss": 0.6658,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.459523173976314,
"learning_rate": 3.235576923489527e-05,
"loss": 0.6575,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.49899414949450965,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.6657,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.4866294474961543,
"learning_rate": 3.226662977303494e-05,
"loss": 0.6566,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.4352922846627129,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.6632,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.45671628273138903,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.6615,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.48537695342769666,
"learning_rate": 3.21321858487569e-05,
"loss": 0.6823,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.48045662214106155,
"learning_rate": 3.208717671149931e-05,
"loss": 0.6762,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.5144523944252073,
"learning_rate": 3.204207092549882e-05,
"loss": 0.6538,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.499875946611401,
"learning_rate": 3.19968688514201e-05,
"loss": 0.6736,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.4633145853462881,
"learning_rate": 3.195157085069774e-05,
"loss": 0.6568,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.46646201712235874,
"learning_rate": 3.190617728553332e-05,
"loss": 0.6688,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.4498918379980022,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.6754,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.46402257562712906,
"learning_rate": 3.181510491450254e-05,
"loss": 0.6872,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.39283804482316026,
"learning_rate": 3.176942683684842e-05,
"loss": 0.6818,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.45292843799250426,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.6719,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.45885694587861053,
"learning_rate": 3.167778872346327e-05,
"loss": 0.6717,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.4252767831605492,
"learning_rate": 3.163182942046815e-05,
"loss": 0.6637,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.4986698921356125,
"learning_rate": 3.158577710967494e-05,
"loss": 0.6644,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.4087916389311612,
"learning_rate": 3.153963215931669e-05,
"loss": 0.6753,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.5322593379038604,
"learning_rate": 3.14933949383672e-05,
"loss": 0.6724,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.47470545340363973,
"learning_rate": 3.144706581653805e-05,
"loss": 0.677,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.4536011480338454,
"learning_rate": 3.140064516427566e-05,
"loss": 0.6727,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.5229849799410862,
"learning_rate": 3.135413335275833e-05,
"loss": 0.66,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.4717218600847639,
"learning_rate": 3.130753075389327e-05,
"loss": 0.682,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.4329408484498814,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.6833,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.44855896182645066,
"learning_rate": 3.121405468537547e-05,
"loss": 0.6599,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.44833125439808774,
"learning_rate": 3.116718196315488e-05,
"loss": 0.6669,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.47098843698038906,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.6693,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.41197357991831984,
"learning_rate": 3.107316901675252e-05,
"loss": 0.679,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.49932967167709486,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.6677,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.525052526937997,
"learning_rate": 3.097880190800057e-05,
"loss": 0.6784,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.48680075098880665,
"learning_rate": 3.093148648549788e-05,
"loss": 0.6819,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.4676703523529741,
"learning_rate": 3.088408365512055e-05,
"loss": 0.6626,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.5721297794286264,
"learning_rate": 3.083659379590034e-05,
"loss": 0.6783,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.45421814916621184,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.6575,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.4417312199396767,
"learning_rate": 3.074135451053475e-05,
"loss": 0.6644,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.4217882697271688,
"learning_rate": 3.069360584592015e-05,
"loss": 0.6716,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.4271827814395117,
"learning_rate": 3.064577167551816e-05,
"loss": 0.6829,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.46429332598837186,
"learning_rate": 3.059785238180953e-05,
"loss": 0.677,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.4072355316227741,
"learning_rate": 3.054984834795566e-05,
"loss": 0.67,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.4329964395865425,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.6498,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.44036930818476344,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.6722,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.407373340584056,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.6812,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.4319705074050344,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.6639,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.340211819610919,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.661,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.42114772026000885,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.6872,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.40613695135259587,
"learning_rate": 3.021147971398632e-05,
"loss": 0.6788,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.3859135203362756,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.6496,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.41501189012644263,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.6567,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.41441534104384914,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.6672,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.411653114732816,
"learning_rate": 3.001632138223332e-05,
"loss": 0.6559,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.6592279375652497,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.5611,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 1.0952468845028034,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.5722,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.8691159497836175,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.5879,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.7480094753235262,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.5562,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.679603003993355,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.5704,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.7614516847591147,
"learning_rate": 2.972118904869376e-05,
"loss": 0.5592,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.5120423546696438,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.5561,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.9644304601846198,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.5592,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.5361807985736651,
"learning_rate": 2.95725682050846e-05,
"loss": 0.5632,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.7935648040290094,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.5515,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.5908703832792701,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.5589,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.6351905895736177,
"learning_rate": 2.942325848714505e-05,
"loss": 0.5695,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.5832302654696425,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.5654,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.4960024692089391,
"learning_rate": 2.932334127129686e-05,
"loss": 0.5574,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.5982036531241831,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.5615,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.6640539694803247,
"learning_rate": 2.922312585929131e-05,
"loss": 0.5592,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.45623014864615613,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.5455,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.7866113496523254,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.5558,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.5589654967292341,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.5519,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.7150002513532401,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.5686,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.5073387915941021,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.5423,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.6003358517471923,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.5651,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.49507404048328024,
"learning_rate": 2.887006998351372e-05,
"loss": 0.5691,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.547455776544649,
"learning_rate": 2.881934649605535e-05,
"loss": 0.5623,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.4754753803344939,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.5677,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.4876264152525405,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.5703,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.49123568246155264,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.574,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.396758750941845,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.5498,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.44861087840980113,
"learning_rate": 2.856467940121025e-05,
"loss": 0.5578,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.3913641600938258,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.5611,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.4156654370163869,
"learning_rate": 2.846233032790149e-05,
"loss": 0.547,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.4673175034273253,
"learning_rate": 2.84110540896642e-05,
"loss": 0.5665,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.45549097523398246,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.5552,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.4261136197093578,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.54,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.4830165345637051,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.5597,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.4395661223795321,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.5543,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.4611952241517715,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.5707,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.5256534788044066,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.5735,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.43145073899869646,
"learning_rate": 2.805026032173773e-05,
"loss": 0.5827,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.4989828546143551,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.5514,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.43270359995188623,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.5612,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.4379277967544891,
"learning_rate": 2.7894660620408e-05,
"loss": 0.5598,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.446272309102627,
"learning_rate": 2.784266724883173e-05,
"loss": 0.548,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.47347333791846363,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.5474,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.4660560962366183,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.566,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.5446435773830683,
"learning_rate": 2.768631254100716e-05,
"loss": 0.5788,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.3957162595765825,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.5563,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.5643085680745933,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.5674,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.4737182969277464,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.5499,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.45794869791302145,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.5467,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.4946616425471109,
"learning_rate": 2.742449775134085e-05,
"loss": 0.5639,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.4565998233252808,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.5577,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.49086619932285985,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.5585,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.4749464103989116,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.5767,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.4840012015551207,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.5643,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.449716583784905,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.557,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.4151693969636063,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.554,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.517878109750726,
"learning_rate": 2.705547548999996e-05,
"loss": 0.5677,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.423804046787156,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.5737,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.4827164351444954,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.5593,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.4974514903342469,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.5796,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.46906656201615077,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.5575,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.4763445121627248,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.5618,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.4799844331866313,
"learning_rate": 2.67369650552155e-05,
"loss": 0.5621,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.4367920081888897,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.5672,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.4335627355998808,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.5843,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.48908778908135725,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.5646,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.4215533558311586,
"learning_rate": 2.652354173593614e-05,
"loss": 0.588,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.43563752989064675,
"learning_rate": 2.647005443331317e-05,
"loss": 0.5682,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.40025872970056786,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.566,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.3915090489127152,
"learning_rate": 2.636292505301542e-05,
"loss": 0.5752,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.4370620802917011,
"learning_rate": 2.630928383194431e-05,
"loss": 0.5576,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.4260004163264931,
"learning_rate": 2.625559216200781e-05,
"loss": 0.5624,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.5011257598438569,
"learning_rate": 2.620185047252308e-05,
"loss": 0.5965,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.3968063979111228,
"learning_rate": 2.614805919320723e-05,
"loss": 0.5626,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.4369499206109972,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.5788,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.425312611771855,
"learning_rate": 2.60403295859298e-05,
"loss": 0.558,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.4473354693650285,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.5707,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.42334492208602426,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.5493,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.46694676741662006,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.5704,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.41561174103171306,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.5699,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.505344999596846,
"learning_rate": 2.577016790136453e-05,
"loss": 0.5829,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.38543475793636195,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.5748,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.5338361278450793,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.5875,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.4101278977082694,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.5679,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.4858218081695534,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.5767,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.44279372170476006,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.5779,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.4997946088700214,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.5556,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.4987235618471435,
"learning_rate": 2.539001591449889e-05,
"loss": 0.5688,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.47321325478911835,
"learning_rate": 2.53355326151989e-05,
"loss": 0.5566,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.46278874530518443,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.5736,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.43387941624003395,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.5675,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.4385861882193394,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.5653,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.4447076312572787,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.5874,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.44260749507563335,
"learning_rate": 2.506248490185627e-05,
"loss": 0.5582,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.40370577091343457,
"learning_rate": 2.500775217241923e-05,
"loss": 0.5828,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.47083851071126953,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.5823,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.41441044165089963,
"learning_rate": 2.4898167025948e-05,
"loss": 0.5677,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.5120353513675325,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.5815,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.44419002257198326,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.587,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.4882459172690997,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.564,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.3892982546232286,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.5619,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.5543916333864624,
"learning_rate": 2.46235264422994e-05,
"loss": 0.566,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.4235034927881744,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.5802,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.4643153198863151,
"learning_rate": 2.451340834511592e-05,
"loss": 0.5639,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.4336570888848717,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.585,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.44699470387458273,
"learning_rate": 2.440314589184958e-05,
"loss": 0.5634,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.43234732955018895,
"learning_rate": 2.434796163357541e-05,
"loss": 0.5858,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.38911027743709037,
"learning_rate": 2.429274260911588e-05,
"loss": 0.5606,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.43850328271260086,
"learning_rate": 2.423748926000084e-05,
"loss": 0.5796,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.4122377604034178,
"learning_rate": 2.418220202803459e-05,
"loss": 0.5828,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.41853010749620106,
"learning_rate": 2.412688135529237e-05,
"loss": 0.5704,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.43927984133326653,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.562,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.42610080112214094,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.5735,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.39510282270684804,
"learning_rate": 2.396072311715187e-05,
"loss": 0.5603,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.49712079817345184,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.5772,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.37791243172162764,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.5759,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.46752286188722714,
"learning_rate": 2.379427985200223e-05,
"loss": 0.5839,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.39828478534402495,
"learning_rate": 2.373873749394972e-05,
"loss": 0.567,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.42995616282590565,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.573,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.3963679677693732,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.5775,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.4385464354757159,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.5616,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.40759604534530963,
"learning_rate": 2.351627355800233e-05,
"loss": 0.5745,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.3812923584277161,
"learning_rate": 2.346058617160256e-05,
"loss": 0.5683,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.42939007478864755,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.5703,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.4136101717575995,
"learning_rate": 2.334912883203469e-05,
"loss": 0.5742,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.42349141823310615,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.5754,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.3833583575204637,
"learning_rate": 2.323756437450136e-05,
"loss": 0.5582,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.42777549025071754,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.5702,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.41045689373132677,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.5691,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.38149201732639676,
"learning_rate": 2.307002464849052e-05,
"loss": 0.563,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.4083429669658198,
"learning_rate": 2.30141283818846e-05,
"loss": 0.5523,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.4169308634517512,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.5849,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.3902529696786323,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.586,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.40672128029571203,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.5823,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.39021661094974736,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.577,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.4124677689954698,
"learning_rate": 2.273429448107257e-05,
"loss": 0.5628,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.4592125318313731,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.5783,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.38459165020657726,
"learning_rate": 2.262220474347102e-05,
"loss": 0.5703,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.44616846643392777,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.5982,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.4095940321371064,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.5582,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.4109709652399218,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.5722,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.47609815653935705,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.5829,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.40319186780683863,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.5489,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.4601462820731242,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.5774,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.4288843962311107,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.5786,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.4285901162671096,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.5775,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.4710112803576252,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.5591,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.4109482849640052,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.5714,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.4366782943159167,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.5713,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.4400591671212513,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.5628,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.3752752234541566,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.5606,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.44246648087593726,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.5689,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.3971836955251671,
"learning_rate": 2.17791064266051e-05,
"loss": 0.5824,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3829335565921508,
"learning_rate": 2.172276926861741e-05,
"loss": 0.5673,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.4083866660552685,
"learning_rate": 2.166641833541255e-05,
"loss": 0.5634,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.38608278193374973,
"learning_rate": 2.161005407757109e-05,
"loss": 0.5596,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.38462131496533297,
"learning_rate": 2.155367694578013e-05,
"loss": 0.5644,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.4071758982509934,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.5837,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.396518830444739,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.5595,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.3637200433640914,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.5418,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.4149193513997958,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.5785,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.4119445032790421,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.5558,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.4214100631305422,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.567,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.3809061109480511,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.5555,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.48525610397719643,
"learning_rate": 2.110225054402733e-05,
"loss": 0.601,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.3968883079024169,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.5906,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.4891093458234197,
"learning_rate": 2.09892967807724e-05,
"loss": 0.5872,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3685454438769812,
"learning_rate": 2.093280780774835e-05,
"loss": 0.5637,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.45420296831904894,
"learning_rate": 2.087631137601793e-05,
"loss": 0.572,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.47697494859165646,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.5747,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.4488297900175968,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.5689,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.37724933161943974,
"learning_rate": 2.070678184630458e-05,
"loss": 0.5657,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.43629389591780104,
"learning_rate": 2.06502600977305e-05,
"loss": 0.5709,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.39524470648418175,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.5576,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.4456391663867885,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.5594,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.3788370034646185,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.5815,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.44545019962605653,
"learning_rate": 2.042412562887594e-05,
"loss": 0.5731,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.3772186650217795,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.5745,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.4669313767236801,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.5809,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.4269293191454068,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.5562,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.41076065483196006,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.5775,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.49215333404880596,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.5644,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.3788615515360649,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.5664,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.45790154639919706,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.5737,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.4098004763257171,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.5755,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.4383917110407997,
"learning_rate": 1.99151687695709e-05,
"loss": 0.5692,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.3639354570127422,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.575,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.4161257980301455,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.572,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.39962075178059225,
"learning_rate": 1.974551241345447e-05,
"loss": 0.5706,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.3724728264467276,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.5787,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.37562601878500956,
"learning_rate": 1.963241759676897e-05,
"loss": 0.5758,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.4427424614637918,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.5586,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.38535451081777594,
"learning_rate": 1.951933453677638e-05,
"loss": 0.5717,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.4217715107851429,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.5511,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.368048541784159,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.5737,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.4114860757826348,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.5548,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.381619100303788,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.5735,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3818092359650817,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.576,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.36530954713778657,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.5722,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.39336824535784304,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.5935,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.35563879648505653,
"learning_rate": 1.906719219225165e-05,
"loss": 0.561,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.3674178002153716,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.5663,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.3799018014947842,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.5741,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.4165102432453501,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.5622,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.38805356561439286,
"learning_rate": 1.884128556891707e-05,
"loss": 0.5897,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.3968222780940193,
"learning_rate": 1.878483094691119e-05,
"loss": 0.5639,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3834215928448922,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.5681,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.38194579649429905,
"learning_rate": 1.867195130360947e-05,
"loss": 0.5482,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.4457447994268672,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.5775,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.38023538502349075,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.5506,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.43060605663081286,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.5678,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.3595842880318931,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.5396,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.41617058362268444,
"learning_rate": 1.838994592242891e-05,
"loss": 0.5898,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.3767470229349702,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.5739,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.4137433691380765,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.5779,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.3886671785919736,
"learning_rate": 1.82208935733949e-05,
"loss": 0.5782,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.4041177307457197,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.5698,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.38915817548493414,
"learning_rate": 1.810826238483892e-05,
"loss": 0.5648,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.37123422524703714,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.5601,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.3923058116900741,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.5506,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.437941468313617,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.5666,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.3747755766426488,
"learning_rate": 1.788318512320838e-05,
"loss": 0.5694,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.37482224519506213,
"learning_rate": 1.782695699829051e-05,
"loss": 0.5782,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3858542662390469,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.558,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.3697663843774481,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.564,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.37442398711829356,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.5584,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.4484918241065211,
"learning_rate": 1.76022227484387e-05,
"loss": 0.5708,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.37790840172391,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.5755,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.40476242468143964,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.5746,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.4141585585195106,
"learning_rate": 1.743387180005809e-05,
"loss": 0.5763,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.3786449374905892,
"learning_rate": 1.737779525652899e-05,
"loss": 0.5707,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.4307060105245239,
"learning_rate": 1.732173968007914e-05,
"loss": 0.5855,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.38229345506543266,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.5845,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.4028070829871976,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.5655,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.39995773232256326,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.569,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.42850647741948844,
"learning_rate": 1.709773600686015e-05,
"loss": 0.5832,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.37369809482820227,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.5909,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.3885666779192199,
"learning_rate": 1.69858716181154e-05,
"loss": 0.5685,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.40371542935910104,
"learning_rate": 1.692997535150948e-05,
"loss": 0.5643,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.37290578634649807,
"learning_rate": 1.687410363273884e-05,
"loss": 0.5691,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.39769177434948155,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.5835,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.3864927197860913,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.5863,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.36542710303879994,
"learning_rate": 1.67066402299235e-05,
"loss": 0.5671,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.3805076648641701,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.5905,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.3588102997686781,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.5858,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.3820554671006271,
"learning_rate": 1.653941382839745e-05,
"loss": 0.5472,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.6930962952908539,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.4549,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.5543432961833293,
"learning_rate": 1.642806717162757e-05,
"loss": 0.4554,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 1.2399344730511936,
"learning_rate": 1.637243646233718e-05,
"loss": 0.4459,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.7856516043914463,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.4613,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.7610706737352125,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.4588,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.7153060021169937,
"learning_rate": 1.620572014799777e-05,
"loss": 0.4396,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.5206721709991748,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.4621,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.771547577919071,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.4362,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.5506064140597104,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.4368,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.6816076359996414,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.4382,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.5316671971679866,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.4463,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.6160584880401327,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.4334,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.5374063203729846,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.4488,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.5763037890000867,
"learning_rate": 1.576251073999917e-05,
"loss": 0.4403,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.5142016936172218,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.4405,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.5012696663686841,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.4644,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.6328755936106206,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.4629,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.4545521253830397,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.45,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.4799730104536855,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.436,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.4981431739331926,
"learning_rate": 1.543151434154755e-05,
"loss": 0.4463,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.5343173889321837,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.4397,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.46360789912019135,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.4457,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.5793590156399374,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.4532,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.45772906863948515,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.4176,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.5396395094046114,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.4526,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.483244158213586,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.4459,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.4877001032630984,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.4498,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.43657644076011964,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.4343,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.41395476250684615,
"learning_rate": 1.493751509814374e-05,
"loss": 0.4275,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.5103069772571637,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.4262,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.428087079095744,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.4442,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.43534714665911217,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.4454,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.4653075586282004,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.4432,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.41920764387974224,
"learning_rate": 1.466446738480111e-05,
"loss": 0.444,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.42679599927859285,
"learning_rate": 1.460998408550112e-05,
"loss": 0.4339,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.43447394280469215,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.4435,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.3930038491974023,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.4397,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.40579836342539655,
"learning_rate": 1.444679451900437e-05,
"loss": 0.4504,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.40629369223765704,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.4426,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.4160128391945634,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.4451,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.40296363255287715,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.4446,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.41374880071731535,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.4413,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.39905785116849013,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.4484,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.3980248511853502,
"learning_rate": 1.412162598317602e-05,
"loss": 0.4522,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.38011724829969645,
"learning_rate": 1.406759321421907e-05,
"loss": 0.4404,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.36823944576522577,
"learning_rate": 1.401360788062871e-05,
"loss": 0.439,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3917768473722188,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.4465,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.3873258078200155,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.4467,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.409406861946073,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.4504,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.38732903150047,
"learning_rate": 1.379814952747693e-05,
"loss": 0.4457,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.4449692023579043,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.4504,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3950896931839253,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.4483,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.41995334740979234,
"learning_rate": 1.363707494698459e-05,
"loss": 0.453,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.43881821798611376,
"learning_rate": 1.358348460369265e-05,
"loss": 0.4349,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.39764205042215284,
"learning_rate": 1.352994556668684e-05,
"loss": 0.4604,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.4039671160860456,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.4381,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.376369466153675,
"learning_rate": 1.342302312350678e-05,
"loss": 0.434,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.39181213895591865,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.4349,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.3889731666738183,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.4373,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.44682744239540506,
"learning_rate": 1.326303494478451e-05,
"loss": 0.4731,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.3972731691162808,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.4435,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.3850982857701653,
"learning_rate": 1.315664479122898e-05,
"loss": 0.4395,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.4296073468146287,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.431,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.38301535797818725,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.4462,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.3903772068716491,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.4501,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.40924251045109666,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.4439,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.4107644402041248,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.434,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.392004167753336,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.4281,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.41114943456960906,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.4459,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.4336806475811975,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.4765,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.3979019983328977,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.4597,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.4105956405988833,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.4337,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.39099409888347475,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.4376,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.3832767201778155,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.4346,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.42170787050678843,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.4569,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.39227573679292804,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.4399,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.3972606182892081,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.4417,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.38071588083545316,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.4345,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.39046061805029886,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.4273,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.3847318268417395,
"learning_rate": 1.220938883250288e-05,
"loss": 0.4566,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.38236775147780644,
"learning_rate": 1.215733275116827e-05,
"loss": 0.4197,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.3663294416692007,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.4422,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.371593258395308,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.434,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.37185068196253696,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.4444,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.3787141829017015,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.4429,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.38802812978858336,
"learning_rate": 1.189800129803139e-05,
"loss": 0.4575,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.37046808137422166,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.4507,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.3815411729781208,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.4305,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.3900084872393736,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.4393,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3748819783043692,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.4489,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.3894633129253501,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.4522,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.39975185993822815,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.463,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.37283460260848206,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.4221,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.36221337893696876,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.458,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.3917686751840278,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.4351,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.38313146806980747,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.4452,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.37227750396355236,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.4489,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.40155006867308923,
"learning_rate": 1.128231163049258e-05,
"loss": 0.4434,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.3762539132408165,
"learning_rate": 1.123144751066259e-05,
"loss": 0.4292,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.3771096634989867,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.4416,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.37942837027439763,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.4452,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.3871404287448122,
"learning_rate": 1.107927745387114e-05,
"loss": 0.4408,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.3794693924644305,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.4545,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.3642817810260331,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.4378,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.39097403682664283,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.4443,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.37352830756570127,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.4358,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.3766208387756165,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.4476,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.37142760668747116,
"learning_rate": 1.077687414070869e-05,
"loss": 0.4369,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.39075504226989377,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.4582,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.3901142412684026,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.449,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.3749466190164667,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.4495,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.36591555202060144,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.4465,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.39174953673692586,
"learning_rate": 1.052689572753247e-05,
"loss": 0.4352,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.41251636927944446,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.4422,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.3652190494220516,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.4616,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.37703493081225026,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.4391,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.42923253016769775,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.4494,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.3822924635504085,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.4587,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.37596082472702924,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.4494,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.39792165007583385,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.4504,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.37384201150851465,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.4352,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.3832498955138027,
"learning_rate": 1.008173988718005e-05,
"loss": 0.4297,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.37003610823063027,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.4398,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3895887902015032,
"learning_rate": 9.983678617766677e-06,
"loss": 0.4369,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.3719420026862606,
"learning_rate": 9.934767922553788e-06,
"loss": 0.453,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.37674066195279277,
"learning_rate": 9.885937708668371e-06,
"loss": 0.4363,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.3965597415627715,
"learning_rate": 9.837188366555524e-06,
"loss": 0.4454,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.38601193086292124,
"learning_rate": 9.788520286013691e-06,
"loss": 0.4569,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.3874303398467139,
"learning_rate": 9.739933856191537e-06,
"loss": 0.4573,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.37597526325899017,
"learning_rate": 9.691429465584866e-06,
"loss": 0.4429,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.39866100988467384,
"learning_rate": 9.643007502033492e-06,
"loss": 0.4345,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.3937701774486641,
"learning_rate": 9.594668352718152e-06,
"loss": 0.4527,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.38581642488382795,
"learning_rate": 9.546412404157383e-06,
"loss": 0.436,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.3798082369137003,
"learning_rate": 9.498240042204471e-06,
"loss": 0.4303,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.3717665650814942,
"learning_rate": 9.450151652044339e-06,
"loss": 0.4231,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.37707058142191363,
"learning_rate": 9.402147618190472e-06,
"loss": 0.4369,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.3671132752580181,
"learning_rate": 9.354228324481841e-06,
"loss": 0.4599,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.38244482959481885,
"learning_rate": 9.306394154079852e-06,
"loss": 0.4367,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.37634087230372143,
"learning_rate": 9.258645489465254e-06,
"loss": 0.4564,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.3784912780066624,
"learning_rate": 9.21098271243509e-06,
"loss": 0.4407,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.3662538043724635,
"learning_rate": 9.163406204099659e-06,
"loss": 0.4389,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.38884736174239387,
"learning_rate": 9.115916344879452e-06,
"loss": 0.4502,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.36887405960434444,
"learning_rate": 9.068513514502126e-06,
"loss": 0.4401,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.36246205000204174,
"learning_rate": 9.021198091999439e-06,
"loss": 0.4677,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.3964046829881989,
"learning_rate": 8.973970455704251e-06,
"loss": 0.4459,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.362258654660145,
"learning_rate": 8.926830983247495e-06,
"loss": 0.436,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.3850623398971165,
"learning_rate": 8.879780051555116e-06,
"loss": 0.4566,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.3772941586990416,
"learning_rate": 8.83281803684512e-06,
"loss": 0.4218,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.362818390020583,
"learning_rate": 8.785945314624532e-06,
"loss": 0.4263,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.3618319092691258,
"learning_rate": 8.739162259686385e-06,
"loss": 0.4397,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.3545117085195587,
"learning_rate": 8.692469246106734e-06,
"loss": 0.4507,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.3655098969467907,
"learning_rate": 8.645866647241677e-06,
"loss": 0.4308,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.36577776430997244,
"learning_rate": 8.599354835724348e-06,
"loss": 0.4601,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.37339895155320085,
"learning_rate": 8.552934183461958e-06,
"loss": 0.4307,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.36181013478043106,
"learning_rate": 8.506605061632805e-06,
"loss": 0.4267,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.36701753983154795,
"learning_rate": 8.460367840683308e-06,
"loss": 0.435,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.3736533720640735,
"learning_rate": 8.414222890325065e-06,
"loss": 0.4414,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3813031943859201,
"learning_rate": 8.368170579531855e-06,
"loss": 0.4412,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3479100698103644,
"learning_rate": 8.322211276536733e-06,
"loss": 0.4406,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.3568953680744935,
"learning_rate": 8.27634534882907e-06,
"loss": 0.4468,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.3589782595663166,
"learning_rate": 8.230573163151585e-06,
"loss": 0.4536,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.35416743151682195,
"learning_rate": 8.184895085497466e-06,
"loss": 0.4314,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.3544216588935654,
"learning_rate": 8.139311481107401e-06,
"loss": 0.4483,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.35350200387191166,
"learning_rate": 8.093822714466686e-06,
"loss": 0.4419,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.35098608157320954,
"learning_rate": 8.048429149302271e-06,
"loss": 0.4262,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.35097019828582343,
"learning_rate": 8.003131148579902e-06,
"loss": 0.4505,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.36078678751300847,
"learning_rate": 7.957929074501182e-06,
"loss": 0.4672,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.3586024981137939,
"learning_rate": 7.912823288500695e-06,
"loss": 0.4525,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.3587897541064069,
"learning_rate": 7.867814151243102e-06,
"loss": 0.4378,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.3475526203735357,
"learning_rate": 7.822902022620267e-06,
"loss": 0.4321,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.34551054329302255,
"learning_rate": 7.778087261748377e-06,
"loss": 0.4368,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.36300748773860464,
"learning_rate": 7.733370226965056e-06,
"loss": 0.4493,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.36298513953415634,
"learning_rate": 7.688751275826527e-06,
"loss": 0.4405,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.3525096504007524,
"learning_rate": 7.644230765104735e-06,
"loss": 0.4287,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.3602324890289102,
"learning_rate": 7.599809050784502e-06,
"loss": 0.4359,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.3499805184726626,
"learning_rate": 7.555486488060661e-06,
"loss": 0.442,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.3513459010325891,
"learning_rate": 7.511263431335249e-06,
"loss": 0.4298,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.35726982244336153,
"learning_rate": 7.467140234214656e-06,
"loss": 0.4522,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.3628392229796561,
"learning_rate": 7.423117249506777e-06,
"loss": 0.4421,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.35698513518113495,
"learning_rate": 7.379194829218231e-06,
"loss": 0.4381,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.3418892545987886,
"learning_rate": 7.335373324551518e-06,
"loss": 0.4412,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.35254094056311425,
"learning_rate": 7.291653085902224e-06,
"loss": 0.4478,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.36073929770048174,
"learning_rate": 7.248034462856213e-06,
"loss": 0.4593,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.36758882832659595,
"learning_rate": 7.20451780418683e-06,
"loss": 0.4443,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.35223732168274624,
"learning_rate": 7.161103457852128e-06,
"loss": 0.4441,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.3655195723592302,
"learning_rate": 7.117791770992046e-06,
"loss": 0.4443,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.3481863140384067,
"learning_rate": 7.074583089925688e-06,
"loss": 0.4375,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.35949230028138696,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.4471,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.36896022674143686,
"learning_rate": 6.988476126329604e-06,
"loss": 0.4382,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.3506373138262087,
"learning_rate": 6.945578532308857e-06,
"loss": 0.4321,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.3684123740773038,
"learning_rate": 6.902785321094301e-06,
"loss": 0.4394,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.35633600933768406,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.453,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.3596746599615412,
"learning_rate": 6.817513414939892e-06,
"loss": 0.4405,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.36881492234108987,
"learning_rate": 6.775035401831924e-06,
"loss": 0.4273,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.35984835478382443,
"learning_rate": 6.732663135188473e-06,
"loss": 0.4423,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.3629461175636608,
"learning_rate": 6.690396953817055e-06,
"loss": 0.4439,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.35406481333721873,
"learning_rate": 6.648237195676927e-06,
"loss": 0.4294,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.3718345678756416,
"learning_rate": 6.606184197876395e-06,
"loss": 0.4518,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.36890206069285947,
"learning_rate": 6.564238296670109e-06,
"loss": 0.4481,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.34718959829122065,
"learning_rate": 6.52239982745638e-06,
"loss": 0.4355,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.3629482564083741,
"learning_rate": 6.480669124774484e-06,
"loss": 0.445,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3639742238947038,
"learning_rate": 6.439046522302013e-06,
"loss": 0.4272,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.3626168260695552,
"learning_rate": 6.397532352852191e-06,
"loss": 0.4433,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.3757676231225775,
"learning_rate": 6.356126948371202e-06,
"loss": 0.4469,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.35337046079366774,
"learning_rate": 6.314830639935559e-06,
"loss": 0.4453,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.3549004759156057,
"learning_rate": 6.273643757749448e-06,
"loss": 0.4484,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.35177672497209284,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.4374,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.3765002657934742,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.4488,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.36446594332688126,
"learning_rate": 6.150742957589784e-06,
"loss": 0.4458,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.35897170213891627,
"learning_rate": 6.109997064904774e-06,
"loss": 0.4424,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.36262747410890855,
"learning_rate": 6.069362236313092e-06,
"loss": 0.4379,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.3541906566031016,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.4431,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.37263170013322555,
"learning_rate": 5.988427070179108e-06,
"loss": 0.4518,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.3781470081103097,
"learning_rate": 5.948127379792234e-06,
"loss": 0.4506,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.3541796398816571,
"learning_rate": 5.90794004780439e-06,
"loss": 0.4417,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.3505160204517124,
"learning_rate": 5.867865395552405e-06,
"loss": 0.4414,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.3637174728275402,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.4272,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.36331427923364995,
"learning_rate": 5.788055411095892e-06,
"loss": 0.442,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.35657368875856194,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.4452,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.3583827247482581,
"learning_rate": 5.708699979051613e-06,
"loss": 0.4569,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.34707444287477773,
"learning_rate": 5.66919351390751e-06,
"loss": 0.4215,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.35357887463499993,
"learning_rate": 5.629801637510188e-06,
"loss": 0.4542,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.3560263629090476,
"learning_rate": 5.590524664836043e-06,
"loss": 0.4233,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.35739028701157755,
"learning_rate": 5.551362909942708e-06,
"loss": 0.4497,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.3527284333053749,
"learning_rate": 5.51231668596653e-06,
"loss": 0.455,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.3550626354927505,
"learning_rate": 5.47338630512007e-06,
"loss": 0.4412,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.35556211481196837,
"learning_rate": 5.434572078689617e-06,
"loss": 0.4381,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.3469875655625784,
"learning_rate": 5.395874317032694e-06,
"loss": 0.4276,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.34895456434476574,
"learning_rate": 5.357293329575573e-06,
"loss": 0.4349,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.3486238351059909,
"learning_rate": 5.318829424810797e-06,
"loss": 0.4501,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.34959121864264237,
"learning_rate": 5.280482910294728e-06,
"loss": 0.4389,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.366657913682901,
"learning_rate": 5.242254092645078e-06,
"loss": 0.4591,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.34358245545417304,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.4253,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.3505205511571937,
"learning_rate": 5.166150769707892e-06,
"loss": 0.4428,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.3577121897307196,
"learning_rate": 5.128276872940503e-06,
"loss": 0.4314,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.35067551396815566,
"learning_rate": 5.090521890074953e-06,
"loss": 0.4375,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.3589874623130995,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.4397,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.3595087322033355,
"learning_rate": 5.015369872647442e-06,
"loss": 0.4503,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.355253751125171,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.4329,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.3552084873882865,
"learning_rate": 4.940697121074675e-06,
"loss": 0.4448,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.34843204873576605,
"learning_rate": 4.903541216934864e-06,
"loss": 0.4314,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.3579606658015637,
"learning_rate": 4.8665060236772e-06,
"loss": 0.4469,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.3707997484661848,
"learning_rate": 4.829591837434102e-06,
"loss": 0.4278,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.34488506733259106,
"learning_rate": 4.792798953370424e-06,
"loss": 0.4336,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.36141504333307367,
"learning_rate": 4.756127665681083e-06,
"loss": 0.4498,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.3523451010836625,
"learning_rate": 4.719578267588725e-06,
"loss": 0.445,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.3460581176667144,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.4306,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.7592175286716759,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.3556,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.5918781101172569,
"learning_rate": 4.610664328486485e-06,
"loss": 0.3395,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.44327387400171825,
"learning_rate": 4.5746054014809e-06,
"loss": 0.3564,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.6983705516169947,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.3395,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.8967121669070789,
"learning_rate": 4.502857857942204e-06,
"loss": 0.3538,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.6855140519067031,
"learning_rate": 4.467169815100549e-06,
"loss": 0.3283,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.5703203449300145,
"learning_rate": 4.431605972355137e-06,
"loss": 0.3381,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.6703836815095271,
"learning_rate": 4.396166614073516e-06,
"loss": 0.3381,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.6251213128642765,
"learning_rate": 4.360852023627846e-06,
"loss": 0.3317,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.5867417732077193,
"learning_rate": 4.325662483392656e-06,
"loss": 0.334,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.5134203460469129,
"learning_rate": 4.29059827474257e-06,
"loss": 0.345,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.479682937619437,
"learning_rate": 4.255659678050063e-06,
"loss": 0.3512,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.5217984094312895,
"learning_rate": 4.22084697268323e-06,
"loss": 0.3296,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.5504627673105765,
"learning_rate": 4.186160437003535e-06,
"loss": 0.3223,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.5142124876247952,
"learning_rate": 4.151600348363582e-06,
"loss": 0.3395,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.44482951614244376,
"learning_rate": 4.11716698310493e-06,
"loss": 0.359,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.41044035547598257,
"learning_rate": 4.082860616555848e-06,
"loss": 0.3291,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.46499442535749563,
"learning_rate": 4.048681523029125e-06,
"loss": 0.3365,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.49844132714708717,
"learning_rate": 4.014629975819882e-06,
"loss": 0.344,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.4520744813800103,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.3459,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.404748573790288,
"learning_rate": 3.94691060843287e-06,
"loss": 0.3309,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.4453502259657894,
"learning_rate": 3.913243329737353e-06,
"loss": 0.3511,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.4498533602748838,
"learning_rate": 3.879704680319503e-06,
"loss": 0.331,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.44606354885821053,
"learning_rate": 3.846294928353471e-06,
"loss": 0.3466,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.4075194726004992,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.337,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.3869430768218614,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.3343,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.42645281569938237,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.3332,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.41410203197914136,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.3422,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.4023107626985011,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.3371,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.3957293190384596,
"learning_rate": 3.648558150665886e-06,
"loss": 0.3364,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.410105513360391,
"learning_rate": 3.616058102895075e-06,
"loss": 0.3356,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.4053722437031172,
"learning_rate": 3.583689060687079e-06,
"loss": 0.34,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.38939660305637125,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.3341,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.39489561011659385,
"learning_rate": 3.51934502719796e-06,
"loss": 0.3409,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.41935718088930446,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.3464,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.39721742055156245,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.3229,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.3834623922206586,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.3414,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.3915629136102313,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.3411,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.38256589213132186,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.343,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.38200837567536194,
"learning_rate": 3.32948376099375e-06,
"loss": 0.3339,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.3803688354989537,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.3326,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.38704983502120655,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.3408,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.382244614042945,
"learning_rate": 3.236349233544438e-06,
"loss": 0.3495,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.37847455205832436,
"learning_rate": 3.205572144561604e-06,
"loss": 0.3311,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.38939020466258284,
"learning_rate": 3.174929343376374e-06,
"loss": 0.3432,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.3889477651758245,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.3599,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.37629531412873257,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.3336,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.385652066526086,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.3323,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.38036460172821235,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.3325,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.37628346974608284,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.3386,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.3693377704111094,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.3237,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.3827574046978805,
"learning_rate": 2.964210239224696e-06,
"loss": 0.3356,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.3742284735894895,
"learning_rate": 2.934650470499869e-06,
"loss": 0.332,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.3833819985724725,
"learning_rate": 2.905227155855128e-06,
"loss": 0.3433,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.3805430975722414,
"learning_rate": 2.875940530558519e-06,
"loss": 0.3373,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.3704460390086701,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.3391,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.37950973706080116,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.3483,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.3674198146515571,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.334,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.3689419039372126,
"learning_rate": 2.760165589921184e-06,
"loss": 0.3326,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.3789489233399394,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.3306,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.36619045512887566,
"learning_rate": 2.703104292151164e-06,
"loss": 0.3324,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.3692370022980923,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.3341,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.3823836399722678,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.3313,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.3777288537204527,
"learning_rate": 2.618550200289303e-06,
"loss": 0.335,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.36575946343694826,
"learning_rate": 2.59064316684412e-06,
"loss": 0.3213,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.3746426079435682,
"learning_rate": 2.562875338152264e-06,
"loss": 0.3209,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.36715988457514803,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.3291,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.37085317424587894,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.3419,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.3667583419082207,
"learning_rate": 2.480409295328803e-06,
"loss": 0.3433,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.36697052017133014,
"learning_rate": 2.45320049480108e-06,
"loss": 0.3232,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.36919375032682994,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.3506,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.3553615724104821,
"learning_rate": 2.399204021408421e-06,
"loss": 0.3266,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.3710029715547377,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.337,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.3777038396597716,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.3377,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.3848934028971819,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.339,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.36322164527864037,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.3192,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.3642752635953664,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.3325,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.3811453651836541,
"learning_rate": 2.24059906434998e-06,
"loss": 0.333,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.3931559298481117,
"learning_rate": 2.214660694699091e-06,
"loss": 0.3472,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.37360165126525385,
"learning_rate": 2.188864536147355e-06,
"loss": 0.3288,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.36381332677252043,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.331,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.36112970782157877,
"learning_rate": 2.137699676264171e-06,
"loss": 0.3304,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.36462006337998515,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.342,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.3767008429616258,
"learning_rate": 2.08710612114857e-06,
"loss": 0.3349,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.3750937866091113,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.319,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.3636299404613835,
"learning_rate": 2.0370854889762e-06,
"loss": 0.3281,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.37653621509886603,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.3331,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.3659383370795353,
"learning_rate": 1.987639379598427e-06,
"loss": 0.3325,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.37269137607388636,
"learning_rate": 1.963132265887919e-06,
"loss": 0.3384,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.3698712643658553,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.3377,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.3760714115204755,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.3354,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.3756688495176629,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.3353,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.3687203334857024,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.3378,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.3764441695747236,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.3407,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.36807125077270003,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.3322,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.37463849942986827,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.3346,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.37959534267792,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.337,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.37136060464146525,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.3256,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.3720656774384499,
"learning_rate": 1.726025117089718e-06,
"loss": 0.3318,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.3603900086152994,
"learning_rate": 1.70311497206598e-06,
"loss": 0.3337,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.36576736589051,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.3292,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.3672335719263267,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.3379,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.3746614193799447,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.3264,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.3731482506622593,
"learning_rate": 1.612939220355272e-06,
"loss": 0.3225,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.3716960413765872,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.338,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.36638355126057015,
"learning_rate": 1.568732761074121e-06,
"loss": 0.3379,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.3823707669747751,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.3542,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.37281918149851073,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.3486,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.35774474986610527,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.3323,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.3656781474752067,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.3353,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.37276314524829235,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.3324,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.36447913232894796,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.3408,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.3702024220460633,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.3469,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.37009062433870227,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.3299,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.36916736858680194,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.352,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.36166003368645283,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.3232,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.37085318839686526,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.3443,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.36505409459801075,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.3285,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.3659705152148927,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.34,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.3741017321633528,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.3342,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.3769466153870274,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.3334,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.3748042823708724,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.3252,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.3642802207699189,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.3344,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.37132458135939606,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.3376,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.36383017323388295,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.3243,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.36440489347163896,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.3263,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.3671392690402538,
"learning_rate": 1.140415161609505e-06,
"loss": 0.3329,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.3713312810235521,
"learning_rate": 1.121666393356604e-06,
"loss": 0.3383,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.36092172058558614,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.3268,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.36540114652924616,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.3343,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.36705667913870316,
"learning_rate": 1.066326386135097e-06,
"loss": 0.3369,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.37273727416039565,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.33,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.3699661428504118,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.3335,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.3686020752270427,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.3422,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.3656319274725089,
"learning_rate": 9.9465987684501e-07,
"loss": 0.3401,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.3673050602536692,
"learning_rate": 9.771228098856534e-07,
"loss": 0.3334,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.37021213096031824,
"learning_rate": 9.597378493454013e-07,
"loss": 0.3423,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.3661952527343085,
"learning_rate": 9.425051342339353e-07,
"loss": 0.3248,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.3726344419771718,
"learning_rate": 9.254248023435864e-07,
"loss": 0.3327,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.36508857884479357,
"learning_rate": 9.084969902482288e-07,
"loss": 0.3301,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.3673842139006806,
"learning_rate": 8.917218333021993e-07,
"loss": 0.3312,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.36345097859287795,
"learning_rate": 8.750994656392109e-07,
"loss": 0.3273,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.37369290322870785,
"learning_rate": 8.58630020171265e-07,
"loss": 0.339,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.36592897328533847,
"learning_rate": 8.423136285876099e-07,
"loss": 0.3335,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.3597326945922512,
"learning_rate": 8.26150421353682e-07,
"loss": 0.3245,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.36326593908683075,
"learning_rate": 8.101405277100549e-07,
"loss": 0.3346,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.36684911980002605,
"learning_rate": 7.942840756714077e-07,
"loss": 0.3335,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.38101279084056694,
"learning_rate": 7.785811920255093e-07,
"loss": 0.3448,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.3661925152103312,
"learning_rate": 7.630320023322024e-07,
"loss": 0.3333,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.36813559893869713,
"learning_rate": 7.476366309223903e-07,
"loss": 0.3527,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.36356709973118195,
"learning_rate": 7.323952008970537e-07,
"loss": 0.3346,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.36344036858784423,
"learning_rate": 7.173078341262618e-07,
"loss": 0.3347,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.36917715705569865,
"learning_rate": 7.023746512481989e-07,
"loss": 0.3441,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.37076344294274705,
"learning_rate": 6.875957716681902e-07,
"loss": 0.3382,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.3617612369460336,
"learning_rate": 6.729713135577687e-07,
"loss": 0.3384,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.3695134071142646,
"learning_rate": 6.585013938537143e-07,
"loss": 0.3483,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.370029186802123,
"learning_rate": 6.441861282571115e-07,
"loss": 0.3428,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.36303453974831984,
"learning_rate": 6.300256312324471e-07,
"loss": 0.346,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.3620753333337273,
"learning_rate": 6.160200160066776e-07,
"loss": 0.3382,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.3629483429012679,
"learning_rate": 6.021693945683327e-07,
"loss": 0.3529,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.37328830455029743,
"learning_rate": 5.884738776666044e-07,
"loss": 0.3341,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.360376486019537,
"learning_rate": 5.749335748104878e-07,
"loss": 0.3388,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.36685784596463733,
"learning_rate": 5.615485942678778e-07,
"loss": 0.3323,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.3636716736769644,
"learning_rate": 5.483190430647289e-07,
"loss": 0.3321,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.3678320798901037,
"learning_rate": 5.352450269841747e-07,
"loss": 0.3298,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.36212733226055893,
"learning_rate": 5.22326650565701e-07,
"loss": 0.3289,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.37522320096853173,
"learning_rate": 5.095640171043048e-07,
"loss": 0.3483,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.36374544931847647,
"learning_rate": 4.969572286496594e-07,
"loss": 0.3318,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.37049405179017236,
"learning_rate": 4.845063860053079e-07,
"loss": 0.339,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.3645793917976824,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.3375,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.36158396371365575,
"learning_rate": 4.600729351261857e-07,
"loss": 0.3266,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.3599650955766126,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.3421,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.3694868230155944,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.3394,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.3589661252757757,
"learning_rate": 4.245948007320588e-07,
"loss": 0.3325,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.3670806382565794,
"learning_rate": 4.130816799401971e-07,
"loss": 0.3382,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.35154248404095106,
"learning_rate": 4.017251756252827e-07,
"loss": 0.3264,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.3691086289297582,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.3322,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.3660528194796655,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.3363,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.375443923368551,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.3392,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.35836343076943683,
"learning_rate": 3.578671204615347e-07,
"loss": 0.3317,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.3631606121616418,
"learning_rate": 3.472950355547267e-07,
"loss": 0.3317,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.3678287180114134,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.3319,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.35571121868534983,
"learning_rate": 3.266223765343224e-07,
"loss": 0.3236,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.3775101343623268,
"learning_rate": 3.16521967718757e-07,
"loss": 0.3457,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.37100801426662045,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.3413,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.3639496272496256,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.3353,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.354968414104115,
"learning_rate": 2.871653894244597e-07,
"loss": 0.3322,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.361108708763102,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.3247,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.35920871957003225,
"learning_rate": 2.683823233205485e-07,
"loss": 0.3317,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.36665737242489893,
"learning_rate": 2.592274132703287e-07,
"loss": 0.336,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.36471693875465966,
"learning_rate": 2.502303499116754e-07,
"loss": 0.3294,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.36768112621350796,
"learning_rate": 2.413912051848666e-07,
"loss": 0.3312,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.3647351954315134,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.3436,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.3730398852501412,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.3381,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.3586029661313256,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.33,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.36538713506116066,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.3336,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.3681980311830722,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.3384,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.3723354363769434,
"learning_rate": 1.916764975761254e-07,
"loss": 0.3435,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.3530428902390965,
"learning_rate": 1.839446914621923e-07,
"loss": 0.3285,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.36418179273708334,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.3369,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.3612902637057711,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.3231,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.36203755229005063,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.3336,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.36783812973124885,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.3324,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.3599162938311411,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.3346,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.35911476343193743,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.323,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.3677046030753514,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.336,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.3668104133038177,
"learning_rate": 1.277993209368744e-07,
"loss": 0.3516,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.3652086419073081,
"learning_rate": 1.214956261808542e-07,
"loss": 0.3381,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.3634884407375713,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.3349,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.3680716414104921,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.338,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.36183102245091486,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.3269,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.3611914198544706,
"learning_rate": 9.787081204466253e-08,
"loss": 0.3329,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.3574320263169255,
"learning_rate": 9.236233586200627e-08,
"loss": 0.3335,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.3697552128612791,
"learning_rate": 8.70130406196279e-08,
"loss": 0.345,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.3588664318361923,
"learning_rate": 8.182296909035047e-08,
"loss": 0.3408,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.36335633104814463,
"learning_rate": 7.679216277384439e-08,
"loss": 0.3369,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.3649359562261993,
"learning_rate": 7.19206618963053e-08,
"loss": 0.3635,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.3824165747621412,
"learning_rate": 6.720850541012347e-08,
"loss": 0.3499,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.35503645587452837,
"learning_rate": 6.265573099357714e-08,
"loss": 0.3283,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.35837566115301067,
"learning_rate": 5.826237505053067e-08,
"loss": 0.3336,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.36264707364241977,
"learning_rate": 5.4028472710137e-08,
"loss": 0.3379,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.3556272309313933,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.3187,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.3619917718025879,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.3411,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.36194411713348773,
"learning_rate": 4.228381947004812e-08,
"loss": 0.3317,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.36068937554557406,
"learning_rate": 3.868805732811343e-08,
"loss": 0.33,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.3552646262586702,
"learning_rate": 3.525190530455636e-08,
"loss": 0.3355,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.3677667671117117,
"learning_rate": 3.197539087475887e-08,
"loss": 0.3394,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.3598018752732349,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.3529,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.3687909679838782,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.3289,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.36971365903054626,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.342,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.3662230869273659,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.3327,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.363657383015254,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.3464,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.35727381814466525,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.3281,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.367344659458201,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.3407,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.3644998390412262,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.3432,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.3633170193446537,
"learning_rate": 9.67435416023843e-09,
"loss": 0.3328,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.3643662658740342,
"learning_rate": 7.995445897566178e-09,
"loss": 0.3229,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.3716494212199529,
"learning_rate": 6.476393170578021e-09,
"loss": 0.339,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.3667665677633131,
"learning_rate": 5.11720812557348e-09,
"loss": 0.3387,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.3623720973810503,
"learning_rate": 3.917901630563492e-09,
"loss": 0.3392,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.36501982935250704,
"learning_rate": 2.878483275170485e-09,
"loss": 0.3379,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.35457594347467325,
"learning_rate": 1.998961370555108e-09,
"loss": 0.3278,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.3618541587648556,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.3361,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.36258767063828334,
"learning_rate": 7.196337656112384e-10,
"loss": 0.3325,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.3598556715672951,
"learning_rate": 3.198382947511292e-10,
"loss": 0.3296,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.36031130864173433,
"learning_rate": 7.995973352770137e-11,
"loss": 0.3497,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.3507371385940964,
"learning_rate": 0.0,
"loss": 0.3227,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1.6351601386649027e+18,
"train_loss": 0.5656346910636917,
"train_runtime": 7027.9189,
"train_samples_per_second": 22.482,
"train_steps_per_second": 0.176
}
],
"logging_steps": 1,
"max_steps": 1235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.6351601386649027e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}