a1_math_automathtext / trainer_state.json
sedrickkeh's picture
End of training
b7b29e4 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": 6.351724706703296,
"learning_rate": 3.2258064516129035e-07,
"loss": 0.8834,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.562933090809133,
"learning_rate": 6.451612903225807e-07,
"loss": 0.9212,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.394330501822805,
"learning_rate": 9.67741935483871e-07,
"loss": 0.8597,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.081964563405212,
"learning_rate": 1.2903225806451614e-06,
"loss": 0.8714,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.125538081208217,
"learning_rate": 1.6129032258064516e-06,
"loss": 0.8904,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 5.584923187100059,
"learning_rate": 1.935483870967742e-06,
"loss": 0.8397,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 4.452475665634315,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.796,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.1559685669436375,
"learning_rate": 2.580645161290323e-06,
"loss": 0.7951,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.490616397021542,
"learning_rate": 2.903225806451613e-06,
"loss": 0.7716,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.2630541460718843,
"learning_rate": 3.225806451612903e-06,
"loss": 0.7688,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.014159970213801,
"learning_rate": 3.548387096774194e-06,
"loss": 0.7492,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.6201420477960244,
"learning_rate": 3.870967741935484e-06,
"loss": 0.7305,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 3.940065413310585,
"learning_rate": 4.193548387096774e-06,
"loss": 0.7311,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 3.9403335659857976,
"learning_rate": 4.516129032258065e-06,
"loss": 0.7278,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.689366155296003,
"learning_rate": 4.838709677419355e-06,
"loss": 0.7154,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.993898105026866,
"learning_rate": 5.161290322580646e-06,
"loss": 0.6594,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.8019040285055974,
"learning_rate": 5.483870967741935e-06,
"loss": 0.6721,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.371777844623746,
"learning_rate": 5.806451612903226e-06,
"loss": 0.6729,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.8883069821692255,
"learning_rate": 6.129032258064517e-06,
"loss": 0.6875,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.6315451861207526,
"learning_rate": 6.451612903225806e-06,
"loss": 0.6734,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.627066408892898,
"learning_rate": 6.774193548387097e-06,
"loss": 0.635,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.6688868299280655,
"learning_rate": 7.096774193548388e-06,
"loss": 0.6215,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.63371507108912,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.6405,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.2383268835757224,
"learning_rate": 7.741935483870968e-06,
"loss": 0.6199,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.0316954991465994,
"learning_rate": 8.064516129032258e-06,
"loss": 0.6437,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.0761396675299868,
"learning_rate": 8.387096774193549e-06,
"loss": 0.5898,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.9954486192623807,
"learning_rate": 8.70967741935484e-06,
"loss": 0.5779,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.7706073669892148,
"learning_rate": 9.03225806451613e-06,
"loss": 0.5829,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.8488974285356122,
"learning_rate": 9.35483870967742e-06,
"loss": 0.5751,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.8843036046043026,
"learning_rate": 9.67741935483871e-06,
"loss": 0.5669,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.7582828460414094,
"learning_rate": 1e-05,
"loss": 0.5953,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.6316151843891895,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.5844,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.6550924579225949,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.5837,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.6398872972899098,
"learning_rate": 1.096774193548387e-05,
"loss": 0.5857,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.6769786946725863,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.5605,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.5417463765588391,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.5757,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.5952531652156343,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.5754,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6906098357862831,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.5592,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.6636841237402326,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.563,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.6321386347448688,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.5793,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.6389784034870906,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.5632,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.6579118599044425,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.5609,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.6066803630245361,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.5528,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.608419152902038,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.5402,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.526605556638071,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.5616,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.5246850903558653,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.5704,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.5280443755286816,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.5278,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.5043268221846728,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.5403,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.5873946418735454,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.5564,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.5897502418523775,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.5226,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.6042241649009616,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.5537,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.555463005652464,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.5713,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.5809022274303646,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.5166,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.5551359230217964,
"learning_rate": 1.741935483870968e-05,
"loss": 0.547,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.5661287121430243,
"learning_rate": 1.774193548387097e-05,
"loss": 0.5418,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.5740694918839165,
"learning_rate": 1.806451612903226e-05,
"loss": 0.552,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.5384669244708673,
"learning_rate": 1.838709677419355e-05,
"loss": 0.5094,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.5311098055584973,
"learning_rate": 1.870967741935484e-05,
"loss": 0.5199,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.4764040449832385,
"learning_rate": 1.903225806451613e-05,
"loss": 0.537,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.5659501378389996,
"learning_rate": 1.935483870967742e-05,
"loss": 0.5478,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.5247024868723122,
"learning_rate": 1.967741935483871e-05,
"loss": 0.5293,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.6710126106806383,
"learning_rate": 2e-05,
"loss": 0.5197,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.6113153318454371,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.5021,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.5652732711848083,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.5316,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.6635409202315184,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.5333,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.5214295984210952,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.5473,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.6289994401522988,
"learning_rate": 2.161290322580645e-05,
"loss": 0.5159,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.49932462222313556,
"learning_rate": 2.193548387096774e-05,
"loss": 0.5464,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.5780014718811078,
"learning_rate": 2.225806451612903e-05,
"loss": 0.5237,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.6226147623176521,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.5134,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.7222450618266787,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.5372,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.6533102761619638,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.5537,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.6371764112140925,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.4918,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.7433574589530576,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.5313,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.8144068911195849,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.5029,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.6528722491896042,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.5192,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.7394199952818035,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.5231,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.5950789783009697,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.5136,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.797818355064567,
"learning_rate": 2.548387096774194e-05,
"loss": 0.5239,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.6011304452871273,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.5344,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.7027654518373676,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.5298,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.6361572938270228,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.5258,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.6539524094171326,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.4927,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.5855733789151948,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.5419,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.6847548019791058,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.5111,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.6766240395323008,
"learning_rate": 2.774193548387097e-05,
"loss": 0.5054,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.5809359547943159,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.5223,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.6207398874999266,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.5121,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.6464059979510546,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.5221,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.6238100419028171,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.5428,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.6020760744562841,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.5153,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.5754142732655908,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.5037,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.5567970088007218,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5173,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.572486005563883,
"learning_rate": 3.032258064516129e-05,
"loss": 0.5255,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.7673088691609318,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.5262,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.8089617331148428,
"learning_rate": 3.096774193548387e-05,
"loss": 0.5248,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.8557452896983069,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.5149,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.706013643054472,
"learning_rate": 3.161290322580645e-05,
"loss": 0.5481,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.598379505521597,
"learning_rate": 3.193548387096774e-05,
"loss": 0.5027,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.6572807965156694,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.5177,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.6921435078681306,
"learning_rate": 3.258064516129033e-05,
"loss": 0.5394,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.7438514544566425,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.5397,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.6133439259824106,
"learning_rate": 3.322580645161291e-05,
"loss": 0.4983,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.6172961357888511,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.5237,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.7464511520644876,
"learning_rate": 3.387096774193549e-05,
"loss": 0.5072,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.6097496976141435,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.511,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.6675466718929456,
"learning_rate": 3.451612903225807e-05,
"loss": 0.5143,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.7166736224416695,
"learning_rate": 3.483870967741936e-05,
"loss": 0.5138,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.516197474930339,
"learning_rate": 3.516129032258065e-05,
"loss": 0.4986,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.5717123144336063,
"learning_rate": 3.548387096774194e-05,
"loss": 0.5127,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 0.6778319364332646,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.5162,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.5850206563636606,
"learning_rate": 3.612903225806452e-05,
"loss": 0.4933,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.6923547330045177,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.5304,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.6833299330829381,
"learning_rate": 3.67741935483871e-05,
"loss": 0.5092,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.7310270041001846,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.5174,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.7137992908094783,
"learning_rate": 3.741935483870968e-05,
"loss": 0.5201,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.653803296397818,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.5045,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.6842833576830064,
"learning_rate": 3.806451612903226e-05,
"loss": 0.5149,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.816215495967424,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.5143,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.5217068672124667,
"learning_rate": 3.870967741935484e-05,
"loss": 0.5192,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.6646494469391804,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.5187,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.6951856639348163,
"learning_rate": 3.935483870967742e-05,
"loss": 0.5151,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.5359633364811349,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.5393,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.6029744274041112,
"learning_rate": 4e-05,
"loss": 0.5376,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.7494445752402011,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.54,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.6539175392037466,
"learning_rate": 3.999968016170525e-05,
"loss": 0.511,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.5444086059565104,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.5094,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.7458144847726844,
"learning_rate": 3.999872065705065e-05,
"loss": 0.5124,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.6212388832527699,
"learning_rate": 3.999800103862945e-05,
"loss": 0.5028,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.6209503711497377,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.5009,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.7481280720583161,
"learning_rate": 3.999608209836944e-05,
"loss": 0.527,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.6008913610800661,
"learning_rate": 3.999488279187443e-05,
"loss": 0.5632,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.6788703427884917,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.5477,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.7442748319127196,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.5374,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.7887115371620675,
"learning_rate": 3.999032564583976e-05,
"loss": 0.5132,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.7969908287403031,
"learning_rate": 3.998848689546591e-05,
"loss": 0.5402,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.6233674742839114,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.4793,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.548632009696359,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.5262,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.5623185895800435,
"learning_rate": 3.998201174509306e-05,
"loss": 0.5149,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.6864436079439379,
"learning_rate": 3.997953378607961e-05,
"loss": 0.5065,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.6477006706546803,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.501,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.776611378656244,
"learning_rate": 3.997409862168454e-05,
"loss": 0.5162,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.46385032920334346,
"learning_rate": 3.997114145976237e-05,
"loss": 0.5018,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.8376728877960896,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.5491,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.645570575642616,
"learning_rate": 3.996474809469545e-05,
"loss": 0.5196,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.6690235587950323,
"learning_rate": 3.996131194267189e-05,
"loss": 0.5243,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.7457780036280578,
"learning_rate": 3.995771618052995e-05,
"loss": 0.5105,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.6302389141158966,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.5097,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.49797654763111726,
"learning_rate": 3.995004594217343e-05,
"loss": 0.509,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.7009209102736353,
"learning_rate": 3.994597152728987e-05,
"loss": 0.528,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.6047942893999411,
"learning_rate": 3.994173762494947e-05,
"loss": 0.5164,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.7628725123455826,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.517,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.7070383960883526,
"learning_rate": 3.993279149458988e-05,
"loss": 0.5437,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.5474984701799074,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.4884,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.7851185046489259,
"learning_rate": 3.992320783722616e-05,
"loss": 0.5039,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.7212856478653541,
"learning_rate": 3.991817703090965e-05,
"loss": 0.4992,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.5462794597229418,
"learning_rate": 3.991298695938038e-05,
"loss": 0.4971,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.7955923705746187,
"learning_rate": 3.9907637664138e-05,
"loss": 0.4962,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.5432463424588232,
"learning_rate": 3.990212918795534e-05,
"loss": 0.4864,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.8273519011697076,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.4865,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.4433828605070193,
"learning_rate": 3.989063487022416e-05,
"loss": 0.4737,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.7515132946181924,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.5093,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.47261724691853657,
"learning_rate": 3.987850437381915e-05,
"loss": 0.5159,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.6432508553398911,
"learning_rate": 3.987220067906313e-05,
"loss": 0.4945,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.5748515588428337,
"learning_rate": 3.986573808672002e-05,
"loss": 0.5048,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.5458695088372197,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.5231,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.5111831486111023,
"learning_rate": 3.985233641724152e-05,
"loss": 0.5006,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.5069318244997476,
"learning_rate": 3.984539744726554e-05,
"loss": 0.514,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.4637666741516366,
"learning_rate": 3.983829979402038e-05,
"loss": 0.5086,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.5065442542735045,
"learning_rate": 3.983104351425871e-05,
"loss": 0.5465,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.5108177878305991,
"learning_rate": 3.982362866600154e-05,
"loss": 0.4924,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.548930782379939,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.509,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.534496740168914,
"learning_rate": 3.980832350242388e-05,
"loss": 0.5058,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.5460609594477748,
"learning_rate": 3.980043330948306e-05,
"loss": 0.4716,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.5466028484115681,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.5301,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.5342439640335971,
"learning_rate": 3.978417801674579e-05,
"loss": 0.5061,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.45369454159038575,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.4972,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.45004151910659,
"learning_rate": 3.976728995023254e-05,
"loss": 0.4935,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.5040348937580492,
"learning_rate": 3.975860879481514e-05,
"loss": 0.5165,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.47527285058156715,
"learning_rate": 3.974976965008833e-05,
"loss": 0.5446,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.4465312147935482,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.5105,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.4603804926543424,
"learning_rate": 3.973161767667946e-05,
"loss": 0.4945,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.4009986660543915,
"learning_rate": 3.972230499314009e-05,
"loss": 0.5032,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.5029016280423306,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.5208,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.5430152311405682,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.5136,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.5014292953080858,
"learning_rate": 3.969342105253098e-05,
"loss": 0.512,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.4477493535215068,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.498,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.53827287737228,
"learning_rate": 3.967337762346568e-05,
"loss": 0.4829,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.45854663305538335,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.529,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.4872497936063446,
"learning_rate": 3.965270496444528e-05,
"loss": 0.5023,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.52223006883761,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.507,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.5953352478624255,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.5074,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.6072381847654823,
"learning_rate": 3.962051762160151e-05,
"loss": 0.485,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.560649842020136,
"learning_rate": 3.960947462140638e-05,
"loss": 0.4977,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.5566140274390501,
"learning_rate": 3.959827482437472e-05,
"loss": 0.484,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.5579430616580731,
"learning_rate": 3.958691832005981e-05,
"loss": 0.5245,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.5141832301590672,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.5089,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.6277187356526602,
"learning_rate": 3.956373555405774e-05,
"loss": 0.5115,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.5970711725049425,
"learning_rate": 3.955190947773937e-05,
"loss": 0.5377,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.6031121315618232,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.5031,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.45352552138483465,
"learning_rate": 3.952778841127214e-05,
"loss": 0.5169,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.5459787398198344,
"learning_rate": 3.95154936139947e-05,
"loss": 0.5142,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.48777862534655314,
"learning_rate": 3.950304277135034e-05,
"loss": 0.5149,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.4821195091095174,
"learning_rate": 3.94904359828957e-05,
"loss": 0.4811,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.5109065764752131,
"learning_rate": 3.94776733494343e-05,
"loss": 0.4982,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.4711969509917959,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.5257,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.5548875914442474,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.5042,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.4688626081964903,
"learning_rate": 3.943845140573213e-05,
"loss": 0.5048,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.5982375885097609,
"learning_rate": 3.942506642518952e-05,
"loss": 0.496,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.541244813757512,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.5077,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.5115606002155595,
"learning_rate": 3.939783060543167e-05,
"loss": 0.5113,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5591504058213124,
"learning_rate": 3.938397998399333e-05,
"loss": 0.5214,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.4217796558219684,
"learning_rate": 3.936997436876756e-05,
"loss": 0.5135,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.5335766498350772,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5077,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.5097106843730664,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.5205,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.4242135191843134,
"learning_rate": 3.932702868644224e-05,
"loss": 0.5184,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.4535741525770329,
"learning_rate": 3.931240422833181e-05,
"loss": 0.4689,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.70365213947996,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.5127,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.47486144200990377,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.5224,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6089087120644388,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.5339,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.5405529584534309,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.4883,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.5144502091911876,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.4996,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.49997207886463185,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.4979,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.4576972900896622,
"learning_rate": 3.92057159243286e-05,
"loss": 0.4825,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.5270219474678155,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.5121,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.4342824243423642,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.5121,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.5645803318990152,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.4806,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.444013213162549,
"learning_rate": 3.914136997982874e-05,
"loss": 0.5094,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.553282643530646,
"learning_rate": 3.912490053436079e-05,
"loss": 0.5127,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.4436930448996126,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5099,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.4323384396450098,
"learning_rate": 3.909150300975178e-05,
"loss": 0.4862,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.41427034894051007,
"learning_rate": 3.907457519765642e-05,
"loss": 0.4972,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.4354775687557458,
"learning_rate": 3.905749486576607e-05,
"loss": 0.4932,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.3991464518987664,
"learning_rate": 3.90402621506546e-05,
"loss": 0.4986,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.42642775433932734,
"learning_rate": 3.902287719011435e-05,
"loss": 0.5003,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.48035482928338563,
"learning_rate": 3.900534012315499e-05,
"loss": 0.4742,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.4018704583167466,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.5002,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.5686293074783814,
"learning_rate": 3.896981023209777e-05,
"loss": 0.5116,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.3700775113207082,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.5211,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.5049736985356861,
"learning_rate": 3.893367361386491e-05,
"loss": 0.4799,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.4577669143728279,
"learning_rate": 3.891537814248417e-05,
"loss": 0.5016,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.44812377005671356,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5209,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.4779011892270232,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.4737,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.4627248243913565,
"learning_rate": 3.885958483839049e-05,
"loss": 0.4777,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.5034499113924977,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.493,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.49993051120063126,
"learning_rate": 3.882163505079171e-05,
"loss": 0.5143,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.4612074470010739,
"learning_rate": 3.880243433489132e-05,
"loss": 0.5209,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.5192563266701791,
"learning_rate": 3.878308327522702e-05,
"loss": 0.4455,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.4720175635314182,
"learning_rate": 3.876358202652936e-05,
"loss": 0.4345,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.5006168912636014,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.4323,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.48489488640947664,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.4259,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.5093588186931203,
"learning_rate": 3.870417871154801e-05,
"loss": 0.4188,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.48236894347124726,
"learning_rate": 3.868407827802192e-05,
"loss": 0.4045,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.4837760740075918,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.4211,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.48349547448299196,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.4112,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.4048418582413915,
"learning_rate": 3.862288124195319e-05,
"loss": 0.4216,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.41476342601836574,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.4198,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.46153519876895216,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.4061,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.40878059601066274,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.4258,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.4945881206174713,
"learning_rate": 3.853920126717703e-05,
"loss": 0.4204,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.41491429206378216,
"learning_rate": 3.851791025492574e-05,
"loss": 0.416,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.4560067493122704,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.423,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.4997966689661459,
"learning_rate": 3.847488419569866e-05,
"loss": 0.4413,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.43769276126988926,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.4136,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.43077275822395666,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.4078,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.45736803180112284,
"learning_rate": 3.840923760917199e-05,
"loss": 0.4203,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.3775314132536621,
"learning_rate": 3.838706077964473e-05,
"loss": 0.398,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.5023128674152759,
"learning_rate": 3.836473692766944e-05,
"loss": 0.4174,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.4887690778956808,
"learning_rate": 3.834226623174705e-05,
"loss": 0.4038,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.47691882524387946,
"learning_rate": 3.831964887155264e-05,
"loss": 0.449,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.4165462383850748,
"learning_rate": 3.829688502793403e-05,
"loss": 0.3903,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.47959087316696475,
"learning_rate": 3.827397488291029e-05,
"loss": 0.4025,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.396869221331589,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.425,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.4844119023598892,
"learning_rate": 3.822771642257145e-05,
"loss": 0.4086,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.49175271178211555,
"learning_rate": 3.820436847713776e-05,
"loss": 0.3925,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.38309144309395116,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.4014,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.4609869727730784,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.392,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.45920135836974274,
"learning_rate": 3.813345202354145e-05,
"loss": 0.4111,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.4930058052197654,
"learning_rate": 3.810952296329563e-05,
"loss": 0.3988,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.4494455661511768,
"learning_rate": 3.808544909978676e-05,
"loss": 0.42,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.4842506795227929,
"learning_rate": 3.806123062550882e-05,
"loss": 0.392,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.4811210435497593,
"learning_rate": 3.803686773411208e-05,
"loss": 0.4224,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.4427318300271397,
"learning_rate": 3.801236062040158e-05,
"loss": 0.422,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.5544241693448969,
"learning_rate": 3.798770948033553e-05,
"loss": 0.3884,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.4098801658966797,
"learning_rate": 3.796291451102381e-05,
"loss": 0.4323,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.4839760528419601,
"learning_rate": 3.793797591072631e-05,
"loss": 0.4035,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.48697208008139947,
"learning_rate": 3.791289387885144e-05,
"loss": 0.424,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.4455839462504645,
"learning_rate": 3.788766861595443e-05,
"loss": 0.3842,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.42205071140179967,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.4269,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.5080292516532647,
"learning_rate": 3.783678920503983e-05,
"loss": 0.4106,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.4006250314716974,
"learning_rate": 3.781113546385265e-05,
"loss": 0.4195,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.5051071633843871,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.4334,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.4641803195017136,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.3955,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.4875250927003641,
"learning_rate": 3.77333205623025e-05,
"loss": 0.4166,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.5916387238832489,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.411,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.5425757089559049,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.4151,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.4905470170672107,
"learning_rate": 3.765422951112704e-05,
"loss": 0.4078,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.4838857961667342,
"learning_rate": 3.762758321970216e-05,
"loss": 0.4113,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.46859242917992083,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.4097,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.4535023428302945,
"learning_rate": 3.757386800198538e-05,
"loss": 0.4095,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.49523358202955436,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.4123,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.4579093589542042,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.4167,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.4861168059932083,
"learning_rate": 3.749224181796305e-05,
"loss": 0.4262,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.45923056723503664,
"learning_rate": 3.746475306375545e-05,
"loss": 0.435,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.475702744671097,
"learning_rate": 3.743712466184774e-05,
"loss": 0.4322,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.45389255833528996,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.39,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.45627077310300107,
"learning_rate": 3.73814497997107e-05,
"loss": 0.4447,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.4656003551864406,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.4284,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.40837185041140567,
"learning_rate": 3.732521901224724e-05,
"loss": 0.3927,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.4082341209771005,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.4126,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.40958291021019305,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.4177,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.38798490108759087,
"learning_rate": 3.723983441007882e-05,
"loss": 0.4086,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.45181333490048325,
"learning_rate": 3.721109687296781e-05,
"loss": 0.4464,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.3846450465797068,
"learning_rate": 3.718222171638484e-05,
"loss": 0.4173,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.44682832858205734,
"learning_rate": 3.715320917121488e-05,
"loss": 0.4212,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.49423605280918803,
"learning_rate": 3.712405946944149e-05,
"loss": 0.4258,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.39114187570043624,
"learning_rate": 3.709477284414488e-05,
"loss": 0.4153,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.48349830023201107,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.4309,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.44611956299009126,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.4297,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.4389553758015622,
"learning_rate": 3.70060937743295e-05,
"loss": 0.433,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.4522407652740642,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.4138,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.4743430305897737,
"learning_rate": 3.694629409915555e-05,
"loss": 0.438,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.42105733298980885,
"learning_rate": 3.691619088858401e-05,
"loss": 0.407,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.3918025894945737,
"learning_rate": 3.68859524166009e-05,
"loss": 0.4329,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.4419452613635008,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.3924,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.4152500778454921,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.4083,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.43226074168406753,
"learning_rate": 3.67944278554384e-05,
"loss": 0.4007,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.43731278519403044,
"learning_rate": 3.676365076645557e-05,
"loss": 0.4326,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.359522263577087,
"learning_rate": 3.673273963576791e-05,
"loss": 0.4113,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.39775049070428287,
"learning_rate": 3.670169471054001e-05,
"loss": 0.4249,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.37295944542270315,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.4074,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.3799077885487924,
"learning_rate": 3.663920447046888e-05,
"loss": 0.3988,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.3795834378147803,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.4003,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.4205804095368786,
"learning_rate": 3.657618204491937e-05,
"loss": 0.4306,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.39000273381940953,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.4204,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.4866408884043448,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.4014,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.39995449725320575,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.4365,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.4635520813421454,
"learning_rate": 3.644854871713611e-05,
"loss": 0.4306,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.4618266429435006,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.4159,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.43938199812431183,
"learning_rate": 3.638394189710493e-05,
"loss": 0.3908,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.47162352695774473,
"learning_rate": 3.635144184933412e-05,
"loss": 0.3977,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.37688746511649,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.4527,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.4436015659314918,
"learning_rate": 3.628604977762752e-05,
"loss": 0.4322,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.4109931740174038,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.4247,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.38729913483392453,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.4252,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.4589310600731949,
"learning_rate": 3.618698565901727e-05,
"loss": 0.4202,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.3539622625388118,
"learning_rate": 3.615370507164654e-05,
"loss": 0.4251,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.43046481070644593,
"learning_rate": 3.61202953196805e-05,
"loss": 0.4227,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.39598620825996333,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.4209,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.43365853154679795,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.4088,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.4493665324929706,
"learning_rate": 3.601929375279662e-05,
"loss": 0.436,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.41652846450348985,
"learning_rate": 3.598537002418012e-05,
"loss": 0.4175,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.4725208766381309,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.4071,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.4316184035938054,
"learning_rate": 3.591713938344416e-05,
"loss": 0.4083,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.5187896025550365,
"learning_rate": 3.588283301689507e-05,
"loss": 0.423,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.3599505901788516,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.423,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.5912288906160825,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.4454,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.36517877942651616,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.4203,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.4794217481773747,
"learning_rate": 3.574434032194994e-05,
"loss": 0.4048,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.43440971623875857,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.4601,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.43573811104180826,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.4154,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.49103196876792016,
"learning_rate": 3.563914797637216e-05,
"loss": 0.4256,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.45503434170271884,
"learning_rate": 3.560383338592649e-05,
"loss": 0.4205,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.41239649167282305,
"learning_rate": 3.556839402764487e-05,
"loss": 0.4043,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.41579483747626345,
"learning_rate": 3.553283018489946e-05,
"loss": 0.4257,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.40574473726231725,
"learning_rate": 3.54971421420578e-05,
"loss": 0.4223,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.4296451427201294,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.4378,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.4077089535201816,
"learning_rate": 3.54253945985191e-05,
"loss": 0.4284,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.402242044833424,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.3933,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.39386443225912143,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.4298,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.4449696644296738,
"learning_rate": 3.531684894865866e-05,
"loss": 0.4225,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.37640604503842817,
"learning_rate": 3.528042173241128e-05,
"loss": 0.4078,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.4470503309703312,
"learning_rate": 3.524387233431892e-05,
"loss": 0.433,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.3664281194401808,
"learning_rate": 3.520720104662958e-05,
"loss": 0.4167,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.46635724283274693,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.4169,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.4155395838668482,
"learning_rate": 3.513349397632281e-05,
"loss": 0.4144,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.481285273265492,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.4268,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.3972962754130324,
"learning_rate": 3.505930287892533e-05,
"loss": 0.413,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.4584405828978417,
"learning_rate": 3.502202656100099e-05,
"loss": 0.3973,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.38179626289926283,
"learning_rate": 3.498463012735256e-05,
"loss": 0.4047,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.4346475712511469,
"learning_rate": 3.494711387700094e-05,
"loss": 0.4058,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.4507738482612585,
"learning_rate": 3.490947810992505e-05,
"loss": 0.4199,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.4154749864691845,
"learning_rate": 3.48717231270595e-05,
"loss": 0.4418,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.5004621230133307,
"learning_rate": 3.483384923029211e-05,
"loss": 0.4312,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.413255801055949,
"learning_rate": 3.479585672246156e-05,
"loss": 0.4398,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.43846912919852815,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.4385,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.455013428216212,
"learning_rate": 3.471951708970527e-05,
"loss": 0.4122,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.4278785578493334,
"learning_rate": 3.468117057518921e-05,
"loss": 0.4124,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.4490645552160817,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.4305,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.41391185304797384,
"learning_rate": 3.460412568296731e-05,
"loss": 0.415,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.42949158027561546,
"learning_rate": 3.456542792131039e-05,
"loss": 0.4204,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.4427247800838953,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.4424,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.44889293157278465,
"learning_rate": 3.448768331403348e-05,
"loss": 0.4462,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.3754037879595889,
"learning_rate": 3.44486370900573e-05,
"loss": 0.3907,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.46247123964133935,
"learning_rate": 3.440947533516396e-05,
"loss": 0.4193,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.40052575365387794,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.465,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.43292005788718235,
"learning_rate": 3.43308064860925e-05,
"loss": 0.4283,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.3900585869344029,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.4255,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.37776826757128346,
"learning_rate": 3.425167928295015e-05,
"loss": 0.4239,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.4739189353410763,
"learning_rate": 3.421194458890411e-05,
"loss": 0.4031,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.37251974396335535,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.442,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.442829970479175,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.4191,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.3634546172303849,
"learning_rate": 3.409205995219561e-05,
"loss": 0.4231,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.4164976889194386,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.4152,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.351504135672175,
"learning_rate": 3.40115729298209e-05,
"loss": 0.4148,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.42804423606598724,
"learning_rate": 3.397116120327025e-05,
"loss": 0.4121,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.42636219524373303,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.417,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.4532258955171712,
"learning_rate": 3.389000293509524e-05,
"loss": 0.4085,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.42487972801350204,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.4203,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.4066985807652989,
"learning_rate": 3.380840041143494e-05,
"loss": 0.4143,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.4544561145554169,
"learning_rate": 3.376743336885792e-05,
"loss": 0.4107,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.3890322084125121,
"learning_rate": 3.372635624225055e-05,
"loss": 0.4198,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.45659709367900214,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.4148,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.36768546328842167,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.4121,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.5234202614253808,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.4336,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.39586065803945586,
"learning_rate": 3.356095347769799e-05,
"loss": 0.4368,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.47688143311666786,
"learning_rate": 3.351933087522552e-05,
"loss": 0.4041,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.4764673244630073,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.4177,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.4496485465907886,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.4236,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.45581444639920204,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.3837,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.4230710890181119,
"learning_rate": 3.335176280432308e-05,
"loss": 0.4264,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.4557458703794748,
"learning_rate": 3.330960304618295e-05,
"loss": 0.4271,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.4041289983566324,
"learning_rate": 3.326733686481153e-05,
"loss": 0.42,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.4215864874919046,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.4326,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.38035266210956437,
"learning_rate": 3.318248658506012e-05,
"loss": 0.4405,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.45140449658261983,
"learning_rate": 3.313990316514069e-05,
"loss": 0.4123,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.39353252589819987,
"learning_rate": 3.309721467890571e-05,
"loss": 0.4165,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.5204358163827767,
"learning_rate": 3.305442146769115e-05,
"loss": 0.4203,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.42173873137642026,
"learning_rate": 3.30115238736704e-05,
"loss": 0.4162,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.431806132867059,
"learning_rate": 3.296852223985148e-05,
"loss": 0.4144,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.4486338170539921,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.42,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.4175944825450578,
"learning_rate": 3.288220822900796e-05,
"loss": 0.4215,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.42053934560197304,
"learning_rate": 3.283889654214788e-05,
"loss": 0.4211,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.38623086769833626,
"learning_rate": 3.279548219581317e-05,
"loss": 0.4167,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.40167277561987463,
"learning_rate": 3.275196553714379e-05,
"loss": 0.4056,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.3577451242844107,
"learning_rate": 3.270834691409778e-05,
"loss": 0.437,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.4237158246885202,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4301,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.37053213086497616,
"learning_rate": 3.262080517078178e-05,
"loss": 0.4223,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.3966350831902648,
"learning_rate": 3.257688275049323e-05,
"loss": 0.4085,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.34949564208760336,
"learning_rate": 3.253285976578535e-05,
"loss": 0.3952,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.3757802962660929,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.4106,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.3813876336552626,
"learning_rate": 3.244451351193935e-05,
"loss": 0.4103,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.40661039946510535,
"learning_rate": 3.240019094921551e-05,
"loss": 0.4151,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.36678388001311185,
"learning_rate": 3.235576923489527e-05,
"loss": 0.4382,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.4249730143509922,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.4449,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.3575678242731323,
"learning_rate": 3.226662977303494e-05,
"loss": 0.4415,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.4005337461035849,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.4199,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.3908857762081958,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.4355,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.4288819642272104,
"learning_rate": 3.21321858487569e-05,
"loss": 0.4432,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.42875084595147833,
"learning_rate": 3.208717671149931e-05,
"loss": 0.4257,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.35333286024288213,
"learning_rate": 3.204207092549882e-05,
"loss": 0.4029,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.440696138275336,
"learning_rate": 3.19968688514201e-05,
"loss": 0.426,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.4140023411246123,
"learning_rate": 3.195157085069774e-05,
"loss": 0.4222,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.35738914715155623,
"learning_rate": 3.190617728553332e-05,
"loss": 0.4057,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.4391624519950241,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.4184,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.345831104708933,
"learning_rate": 3.181510491450254e-05,
"loss": 0.4289,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.44764031154661693,
"learning_rate": 3.176942683684842e-05,
"loss": 0.4251,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.41856354429261533,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.421,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.3626417581343146,
"learning_rate": 3.167778872346327e-05,
"loss": 0.4024,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.3820180746698907,
"learning_rate": 3.163182942046815e-05,
"loss": 0.4151,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.35276876823985626,
"learning_rate": 3.158577710967494e-05,
"loss": 0.4029,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.37128106646896347,
"learning_rate": 3.153963215931669e-05,
"loss": 0.414,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.36023846916521496,
"learning_rate": 3.14933949383672e-05,
"loss": 0.4127,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.3637413060813467,
"learning_rate": 3.144706581653805e-05,
"loss": 0.424,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.3473692006611996,
"learning_rate": 3.140064516427566e-05,
"loss": 0.4085,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.37931759006766363,
"learning_rate": 3.135413335275833e-05,
"loss": 0.445,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.3351161890744758,
"learning_rate": 3.130753075389327e-05,
"loss": 0.4383,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.3501378422938848,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.403,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.4318236561298152,
"learning_rate": 3.121405468537547e-05,
"loss": 0.4431,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.3852028212049434,
"learning_rate": 3.116718196315488e-05,
"loss": 0.403,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.43597984608872253,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.4186,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.3770617170077158,
"learning_rate": 3.107316901675252e-05,
"loss": 0.4406,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.40046153290301073,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.4178,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.3704957541784096,
"learning_rate": 3.097880190800057e-05,
"loss": 0.4282,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.3659333212512278,
"learning_rate": 3.093148648549788e-05,
"loss": 0.4216,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.35336433978651993,
"learning_rate": 3.088408365512055e-05,
"loss": 0.3937,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.39305896682996067,
"learning_rate": 3.083659379590034e-05,
"loss": 0.4152,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.3460106035800424,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.4482,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.397197509133872,
"learning_rate": 3.074135451053475e-05,
"loss": 0.424,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.3603360190266358,
"learning_rate": 3.069360584592015e-05,
"loss": 0.4136,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.38063162041780724,
"learning_rate": 3.064577167551816e-05,
"loss": 0.4027,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.3568831940061296,
"learning_rate": 3.059785238180953e-05,
"loss": 0.4272,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.3569783292590719,
"learning_rate": 3.054984834795566e-05,
"loss": 0.4248,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.3928179664145324,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.4511,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.3943309348654365,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.4216,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.3666810957236805,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.4248,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.39588141716200864,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.4011,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.36535315234623417,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.4186,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.3863688102046143,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.4145,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.36685519996590193,
"learning_rate": 3.021147971398632e-05,
"loss": 0.4198,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.39945055234387195,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.4251,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.4279949639970197,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.4285,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.41432645605494345,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4264,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.4434268265611787,
"learning_rate": 3.001632138223332e-05,
"loss": 0.4221,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.6158789517356149,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.3047,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.5063645812690903,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.2905,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.8308409772127127,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.3104,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.6262109488213045,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.3127,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.5476131902460964,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.2823,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.4039573175001732,
"learning_rate": 2.972118904869376e-05,
"loss": 0.3043,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.49692822337246045,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.3095,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.5110424698428733,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.307,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.45880357804287153,
"learning_rate": 2.95725682050846e-05,
"loss": 0.3266,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.4485461082005231,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.2914,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.44566958217515507,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.2963,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.45534559872384395,
"learning_rate": 2.942325848714505e-05,
"loss": 0.3181,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.4425524625730518,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.3082,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.4265295850947882,
"learning_rate": 2.932334127129686e-05,
"loss": 0.2923,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.4084324245070778,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.3,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.450002529741149,
"learning_rate": 2.922312585929131e-05,
"loss": 0.3232,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.395017609632048,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.3056,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.41300276685591436,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.305,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.38624409129097237,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.2984,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.47482211477157227,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.3275,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.4431955902002269,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.3116,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.3910122864161504,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.3062,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.5213697061896484,
"learning_rate": 2.887006998351372e-05,
"loss": 0.3219,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.46239907020088034,
"learning_rate": 2.881934649605535e-05,
"loss": 0.3307,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.42216948128078413,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.3068,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.439754002348112,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.3175,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.40256545504505664,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.29,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.433596553949257,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.2885,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.4294416679630344,
"learning_rate": 2.856467940121025e-05,
"loss": 0.312,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.40002750519358116,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.2842,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.39468444003868564,
"learning_rate": 2.846233032790149e-05,
"loss": 0.3127,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.37650605129142306,
"learning_rate": 2.84110540896642e-05,
"loss": 0.3074,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.42328153098092175,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.2987,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.35729242794452465,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.2978,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.37875618829656355,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.3022,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.3855909851219303,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.3247,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.37629497480363094,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.3181,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.4538448019928653,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.319,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.349049396563954,
"learning_rate": 2.805026032173773e-05,
"loss": 0.3047,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.39418806834011483,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.2921,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.34575447776284923,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.3084,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.32579021396233604,
"learning_rate": 2.7894660620408e-05,
"loss": 0.3064,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.3723353981379635,
"learning_rate": 2.784266724883173e-05,
"loss": 0.3022,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.3481276866372949,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.2913,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.3711878328565923,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.3104,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.34333836601159873,
"learning_rate": 2.768631254100716e-05,
"loss": 0.301,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.3558696012873851,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.3063,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.397659930207154,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.3114,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.3695917189955577,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.3293,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.4299862784656741,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.3048,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.3952344473666711,
"learning_rate": 2.742449775134085e-05,
"loss": 0.2811,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.3735160308788477,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.3147,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.4044411119702852,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.3175,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.3662333510735184,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.3029,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.3965211235301028,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.3221,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.3698362995695025,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.3228,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.390921301494128,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.3242,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.34922730644824174,
"learning_rate": 2.705547548999996e-05,
"loss": 0.345,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.38247472162551094,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.3294,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.3346936500116944,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.3145,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.3585080905032369,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.3141,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.34922041881806526,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.3129,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.3482801126850834,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.3066,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.3526393985555093,
"learning_rate": 2.67369650552155e-05,
"loss": 0.2904,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.3424319340159419,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.3094,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.38370993655093766,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.3059,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.35350074972824197,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.3256,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.35778400116367237,
"learning_rate": 2.652354173593614e-05,
"loss": 0.308,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.4131830630322393,
"learning_rate": 2.647005443331317e-05,
"loss": 0.3134,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.3369331030812369,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.2949,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.3927416412691185,
"learning_rate": 2.636292505301542e-05,
"loss": 0.2954,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.3507333381196792,
"learning_rate": 2.630928383194431e-05,
"loss": 0.3055,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.386323362871152,
"learning_rate": 2.625559216200781e-05,
"loss": 0.3087,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.4107847272313232,
"learning_rate": 2.620185047252308e-05,
"loss": 0.3209,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.3363937787344964,
"learning_rate": 2.614805919320723e-05,
"loss": 0.2911,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.36220704778777196,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.3089,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.3418460607610858,
"learning_rate": 2.60403295859298e-05,
"loss": 0.3065,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.35442578558492494,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.3378,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.3318504107594776,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.2973,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.36506970274152284,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.3198,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.32980901204025903,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.3097,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.3664718792920613,
"learning_rate": 2.577016790136453e-05,
"loss": 0.3198,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.34322362983082005,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.3285,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.39009223474399074,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.305,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.37375706152844973,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.2965,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.4016658331017777,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.3192,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.3440076358204119,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.3245,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.4192598939248474,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.3091,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.35247887319546845,
"learning_rate": 2.539001591449889e-05,
"loss": 0.3194,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.42533763096994914,
"learning_rate": 2.53355326151989e-05,
"loss": 0.3163,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.3441304335429926,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.2996,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.3846769419794845,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.3209,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.36705498276061566,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.3126,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.37032228263198125,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.3045,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.3705314390848168,
"learning_rate": 2.506248490185627e-05,
"loss": 0.3143,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.3566235613243603,
"learning_rate": 2.500775217241923e-05,
"loss": 0.3077,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.3798704770162515,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.3249,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.38426632557731066,
"learning_rate": 2.4898167025948e-05,
"loss": 0.2965,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.38507315037922696,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.3039,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.35137464840748267,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.3212,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.4143249375047347,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.3199,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.3633960729740527,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.3292,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.3714873029000739,
"learning_rate": 2.46235264422994e-05,
"loss": 0.3246,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.32355774102409157,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.31,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.36788620548946144,
"learning_rate": 2.451340834511592e-05,
"loss": 0.3213,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.37375766564665314,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.3029,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.36297608570217665,
"learning_rate": 2.440314589184958e-05,
"loss": 0.2999,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.36283254564784617,
"learning_rate": 2.434796163357541e-05,
"loss": 0.3133,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.3714691874893284,
"learning_rate": 2.429274260911588e-05,
"loss": 0.3102,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.37084369509429954,
"learning_rate": 2.423748926000084e-05,
"loss": 0.3032,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.3631994560520952,
"learning_rate": 2.418220202803459e-05,
"loss": 0.3263,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.37046832444374006,
"learning_rate": 2.412688135529237e-05,
"loss": 0.2967,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.38073230820334825,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.3262,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.4047279761288131,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.3068,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.3966202197484159,
"learning_rate": 2.396072311715187e-05,
"loss": 0.3115,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.3745951914403979,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.2952,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.38691042202251047,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.3378,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.3908284007426705,
"learning_rate": 2.379427985200223e-05,
"loss": 0.2955,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.37304939581915675,
"learning_rate": 2.373873749394972e-05,
"loss": 0.3325,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.38563285013638593,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.2869,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.4158968591950684,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.3209,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.35991791206976953,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.3106,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.46986561659711584,
"learning_rate": 2.351627355800233e-05,
"loss": 0.3065,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.34840858381429973,
"learning_rate": 2.346058617160256e-05,
"loss": 0.2844,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.4080819149445699,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.2889,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.4034222696148032,
"learning_rate": 2.334912883203469e-05,
"loss": 0.3063,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.3421773714792299,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.3224,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.39747478655666985,
"learning_rate": 2.323756437450136e-05,
"loss": 0.3022,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.3405615746582461,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.3137,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.3777209021668249,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.3198,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.3863561548219596,
"learning_rate": 2.307002464849052e-05,
"loss": 0.3053,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.36419686832322623,
"learning_rate": 2.30141283818846e-05,
"loss": 0.2958,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.3337514550018213,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.3263,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.3902206495510602,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.3245,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.3786531646007775,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.3056,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.3430731968928644,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.3003,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.3561246900767615,
"learning_rate": 2.273429448107257e-05,
"loss": 0.2982,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.34242237504759965,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.2797,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.3874465750657126,
"learning_rate": 2.262220474347102e-05,
"loss": 0.3178,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.35892149869924844,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.3124,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.3795741502713747,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.3315,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.38931625266806436,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.2959,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.3513100524846882,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.3175,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.39630712342173197,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.3357,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.3533191930510549,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.2856,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.3467895277915417,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.3064,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.35182194577166637,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.3029,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.36104960080994936,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.3091,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.3543488560117821,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.3118,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.36893871238997444,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.3022,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.3473386905625485,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.3227,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.4024355132141183,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.315,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.3444556788448552,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.307,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.37456927857737554,
"learning_rate": 2.17791064266051e-05,
"loss": 0.3252,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.33151782544807373,
"learning_rate": 2.172276926861741e-05,
"loss": 0.2901,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.36713595752201034,
"learning_rate": 2.166641833541255e-05,
"loss": 0.3126,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.34669961445136305,
"learning_rate": 2.161005407757109e-05,
"loss": 0.3175,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.33498211912768744,
"learning_rate": 2.155367694578013e-05,
"loss": 0.3093,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.3386045640497428,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.2997,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.35276233986612426,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.3203,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.3340639786879652,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.3035,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.3314888500523006,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.3074,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.3516855122468125,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.3027,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.3302541538442039,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.3206,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.356797698521036,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.2892,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.334856398677841,
"learning_rate": 2.110225054402733e-05,
"loss": 0.307,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.33527265746817897,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.2923,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.37676514992691396,
"learning_rate": 2.09892967807724e-05,
"loss": 0.3183,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.32677187855020146,
"learning_rate": 2.093280780774835e-05,
"loss": 0.2984,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.36003779356290505,
"learning_rate": 2.087631137601793e-05,
"loss": 0.313,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.3502674926568314,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.2969,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.294918444896749,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.3143,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.3560618657585111,
"learning_rate": 2.070678184630458e-05,
"loss": 0.3073,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.3514254595210009,
"learning_rate": 2.06502600977305e-05,
"loss": 0.3225,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.32960457966826384,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.3088,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.339517333876512,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.3297,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.334186257411704,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.3267,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.3288256346453874,
"learning_rate": 2.042412562887594e-05,
"loss": 0.2755,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.3319379718508043,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.3003,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.32987000020892904,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.3073,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.32101099541870964,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.3203,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.3160502845650283,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.3118,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.3190998461041239,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.322,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.32192472707439823,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.3139,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.31198411499034856,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.3022,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.32724219108260716,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.3046,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.3176735595225574,
"learning_rate": 1.99151687695709e-05,
"loss": 0.2929,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.3170515542545952,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.3093,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.3200836099027574,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.3264,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.3491026731901004,
"learning_rate": 1.974551241345447e-05,
"loss": 0.3171,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.3145144897136951,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.2915,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.346386011363863,
"learning_rate": 1.963241759676897e-05,
"loss": 0.2997,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.32453187512177234,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.3138,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.3209713892118004,
"learning_rate": 1.951933453677638e-05,
"loss": 0.3208,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.33526930295884033,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.2975,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.35330151392994336,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.2949,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.31953699266373403,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.3055,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.35821517611036885,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.3176,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3260627241355444,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.3061,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.32109647561583843,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.2849,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.31902872375555114,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.3079,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.3157296467764903,
"learning_rate": 1.906719219225165e-05,
"loss": 0.3198,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.35069678017877765,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.3149,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.34008304046675625,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.3025,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.3176600132320201,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.3267,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.35522765583567645,
"learning_rate": 1.884128556891707e-05,
"loss": 0.2912,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.33956326326852304,
"learning_rate": 1.878483094691119e-05,
"loss": 0.309,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3209556727067354,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.3181,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.3179197427723173,
"learning_rate": 1.867195130360947e-05,
"loss": 0.3057,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.32390712281994133,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.3159,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.3532033404552596,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.3119,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.3048019431210517,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.292,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.358935782391093,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.3064,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.31337773697371435,
"learning_rate": 1.838994592242891e-05,
"loss": 0.3168,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.34777057052969745,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.3163,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.3328011473413515,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.2919,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.32002317850994305,
"learning_rate": 1.82208935733949e-05,
"loss": 0.3269,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.3252334154177748,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.3048,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.3410458203586588,
"learning_rate": 1.810826238483892e-05,
"loss": 0.3083,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.3384460250981647,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.3292,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.359628512074909,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.3126,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.3346653941625629,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.3014,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.33676031362825326,
"learning_rate": 1.788318512320838e-05,
"loss": 0.3226,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.33951856291627297,
"learning_rate": 1.782695699829051e-05,
"loss": 0.2719,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3092697679921745,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.3064,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.3210038565855538,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.2915,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.3174009360716015,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.3258,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.3281400595290637,
"learning_rate": 1.76022227484387e-05,
"loss": 0.325,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.36033994985584294,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.3107,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.3355371706407504,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.322,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.3354143320703414,
"learning_rate": 1.743387180005809e-05,
"loss": 0.3199,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.32075508263529234,
"learning_rate": 1.737779525652899e-05,
"loss": 0.2999,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.32079710636052455,
"learning_rate": 1.732173968007914e-05,
"loss": 0.335,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.3608948689595536,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.3312,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.3091177710273241,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.297,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.34532822409377356,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.3072,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.32234853258101687,
"learning_rate": 1.709773600686015e-05,
"loss": 0.292,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.32601171660312045,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.2914,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.3284486014324784,
"learning_rate": 1.69858716181154e-05,
"loss": 0.2878,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.336252407614942,
"learning_rate": 1.692997535150948e-05,
"loss": 0.2962,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.33871449144977767,
"learning_rate": 1.687410363273884e-05,
"loss": 0.3228,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.3538934620525434,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.3159,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.32475391709678536,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.31,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.3402221670773298,
"learning_rate": 1.67066402299235e-05,
"loss": 0.3007,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.34094132789709725,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.3159,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.3224725417351319,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.286,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.3415048343597573,
"learning_rate": 1.653941382839745e-05,
"loss": 0.3008,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.4807438986558305,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.2256,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.37165337200268733,
"learning_rate": 1.642806717162757e-05,
"loss": 0.2227,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.5794163882363182,
"learning_rate": 1.637243646233718e-05,
"loss": 0.2138,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.5894058220063316,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.2199,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.43252282648482654,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.204,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.4817534718484749,
"learning_rate": 1.620572014799777e-05,
"loss": 0.2359,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.45361022391227296,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.2323,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.39575792067830673,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.22,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.34623048526098893,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.1976,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.47494383835355497,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.2314,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.3569128545915745,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.218,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.3437006757656693,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.2189,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.34929264637511753,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.2212,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.34867620631047397,
"learning_rate": 1.576251073999917e-05,
"loss": 0.2163,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.35468956081042885,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.2155,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.3232871128517303,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.2248,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.34519292139507457,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.1914,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.33352614893276505,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.1938,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.3579007572279721,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.2306,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3309765720353454,
"learning_rate": 1.543151434154755e-05,
"loss": 0.1949,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.31154264745537436,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.2193,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.33596099301118465,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.176,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.35870598191575426,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.2074,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.31990794374284626,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.2013,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.3434285758606235,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.2293,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.34932129034774684,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.2112,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.3337751387637453,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.2124,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.2944635360275809,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.2196,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.33192432308579783,
"learning_rate": 1.493751509814374e-05,
"loss": 0.2186,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.30957834587229743,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.1927,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.32656118488495633,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.2108,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.31562435025477703,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.1986,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.3149996920574632,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.2066,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.32617113294731037,
"learning_rate": 1.466446738480111e-05,
"loss": 0.2139,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.30148310218229907,
"learning_rate": 1.460998408550112e-05,
"loss": 0.2475,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.33566666357074804,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.1995,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.32056491899246037,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.2202,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.3075480819914189,
"learning_rate": 1.444679451900437e-05,
"loss": 0.1763,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.3250931182242424,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.209,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.31620079750854446,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.2149,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.31665396937142193,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.202,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.3042756265473614,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.2045,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.3036852765478824,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.2126,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.3326394296193336,
"learning_rate": 1.412162598317602e-05,
"loss": 0.1785,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.30178990635550906,
"learning_rate": 1.406759321421907e-05,
"loss": 0.191,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.3101214921430777,
"learning_rate": 1.401360788062871e-05,
"loss": 0.2265,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3046630896537524,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.2145,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.3075056551979408,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.2162,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.2934377147012713,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.235,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.2951284198776281,
"learning_rate": 1.379814952747693e-05,
"loss": 0.1954,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.31235401506109234,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.2172,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3042884011014314,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.2126,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.3172906056130839,
"learning_rate": 1.363707494698459e-05,
"loss": 0.1996,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.30793821374529934,
"learning_rate": 1.358348460369265e-05,
"loss": 0.1879,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.2837304537665948,
"learning_rate": 1.352994556668684e-05,
"loss": 0.2358,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.3209703551463685,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.2145,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.2851239376528537,
"learning_rate": 1.342302312350678e-05,
"loss": 0.2186,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.29791946807938596,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.2128,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.30338417386989597,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.2145,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.3076357947233227,
"learning_rate": 1.326303494478451e-05,
"loss": 0.1988,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.3092754337375694,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.2129,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.3153185602035298,
"learning_rate": 1.315664479122898e-05,
"loss": 0.1959,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.2969125158341166,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.2163,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.3207474511776596,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.2166,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.30887888854433115,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.2123,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.29239302249961835,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.2221,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.29707957168489296,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.196,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.29992770444910416,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.2013,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.3028819882174415,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.2226,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.313051949095182,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.2042,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.31944458998075503,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.2262,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.3023774275198997,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.2224,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.3108514664911787,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.2032,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.3191432206982935,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.2156,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.3084519495805069,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.1929,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.29406527755609996,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.214,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.30839830445714683,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.2067,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.3094472665349595,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.2269,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.30553525976989915,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.2093,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.3000473709140589,
"learning_rate": 1.220938883250288e-05,
"loss": 0.2071,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.31071234220548866,
"learning_rate": 1.215733275116827e-05,
"loss": 0.2083,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.31842310382050343,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.2052,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.32038035361514394,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.2205,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.3103241562620045,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.1775,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.3048071412126106,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.2368,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.3202022946583998,
"learning_rate": 1.189800129803139e-05,
"loss": 0.2015,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.3310019434305348,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.2066,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.3393490754269847,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.2115,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.3131235859976247,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.2148,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3120685228945952,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.2185,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.3316616580621037,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.2127,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.32117190805210255,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.2181,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.2860448180655495,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.2306,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.3131534501256737,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.2153,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.3245590909736494,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.2203,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.29308658812009014,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.2222,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.3189764021741067,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.214,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.3366371740112427,
"learning_rate": 1.128231163049258e-05,
"loss": 0.2053,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.3101740700691036,
"learning_rate": 1.123144751066259e-05,
"loss": 0.2129,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.3100523596553614,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.1949,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.31279279625283574,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.202,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.31508089889493324,
"learning_rate": 1.107927745387114e-05,
"loss": 0.2093,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.3076037196957221,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.1991,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.3008245294543836,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.2192,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.29456319252736957,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.2006,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.3017915783079312,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.2303,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.2946615461885256,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.2112,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.31722988555175646,
"learning_rate": 1.077687414070869e-05,
"loss": 0.1903,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.31317542681495253,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.2132,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.30161025463946134,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.2146,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.29703144699707484,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.1974,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.3086031877245761,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.2161,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.2910465132763547,
"learning_rate": 1.052689572753247e-05,
"loss": 0.2125,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.30612543354904376,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.2139,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.30518375348309434,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.2059,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.2981300317313955,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.2155,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.30262761263952953,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.2101,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.28399987782381136,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.2164,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.3160768811673661,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.1986,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.30128889074228454,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.2039,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.29319872382497697,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.2173,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.2903189117249817,
"learning_rate": 1.008173988718005e-05,
"loss": 0.2045,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.28531373298153107,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.2276,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3061989508661014,
"learning_rate": 9.983678617766677e-06,
"loss": 0.2109,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.2997599172052241,
"learning_rate": 9.934767922553788e-06,
"loss": 0.221,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.2901840842074806,
"learning_rate": 9.885937708668371e-06,
"loss": 0.2158,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.2888640326609935,
"learning_rate": 9.837188366555524e-06,
"loss": 0.227,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.2958345414763582,
"learning_rate": 9.788520286013691e-06,
"loss": 0.2023,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.30430026531763477,
"learning_rate": 9.739933856191537e-06,
"loss": 0.203,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.30671490795742856,
"learning_rate": 9.691429465584866e-06,
"loss": 0.1911,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.3046343598394071,
"learning_rate": 9.643007502033492e-06,
"loss": 0.2098,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.3044767214317,
"learning_rate": 9.594668352718152e-06,
"loss": 0.2242,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.2936549202621547,
"learning_rate": 9.546412404157383e-06,
"loss": 0.1984,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.29353421421035725,
"learning_rate": 9.498240042204471e-06,
"loss": 0.2267,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.2999647163874766,
"learning_rate": 9.450151652044339e-06,
"loss": 0.2037,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.3107194495918799,
"learning_rate": 9.402147618190472e-06,
"loss": 0.2318,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.29832622048141516,
"learning_rate": 9.354228324481841e-06,
"loss": 0.2169,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.2872443542486106,
"learning_rate": 9.306394154079852e-06,
"loss": 0.2275,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.31152114125896374,
"learning_rate": 9.258645489465254e-06,
"loss": 0.206,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.31077822639471425,
"learning_rate": 9.21098271243509e-06,
"loss": 0.2105,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.3014254863910834,
"learning_rate": 9.163406204099659e-06,
"loss": 0.2073,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.2900663142149228,
"learning_rate": 9.115916344879452e-06,
"loss": 0.2247,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.302571164891786,
"learning_rate": 9.068513514502126e-06,
"loss": 0.1983,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.28403401089240715,
"learning_rate": 9.021198091999439e-06,
"loss": 0.2041,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.2910301908544143,
"learning_rate": 8.973970455704251e-06,
"loss": 0.2179,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.3110180672267137,
"learning_rate": 8.926830983247495e-06,
"loss": 0.2143,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.3020395901854515,
"learning_rate": 8.879780051555116e-06,
"loss": 0.2004,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.2888949432237277,
"learning_rate": 8.83281803684512e-06,
"loss": 0.2188,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.303973268874925,
"learning_rate": 8.785945314624532e-06,
"loss": 0.2216,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.2891045977517634,
"learning_rate": 8.739162259686385e-06,
"loss": 0.2216,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.28793773001257805,
"learning_rate": 8.692469246106734e-06,
"loss": 0.2102,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.29990129438062096,
"learning_rate": 8.645866647241677e-06,
"loss": 0.2077,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.29964985536206384,
"learning_rate": 8.599354835724348e-06,
"loss": 0.2103,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.2928958328083469,
"learning_rate": 8.552934183461958e-06,
"loss": 0.2192,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.2997484887013067,
"learning_rate": 8.506605061632805e-06,
"loss": 0.2045,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.30957208011578846,
"learning_rate": 8.460367840683308e-06,
"loss": 0.2131,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.30453278311997434,
"learning_rate": 8.414222890325065e-06,
"loss": 0.2006,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.2884925051157935,
"learning_rate": 8.368170579531855e-06,
"loss": 0.2187,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3064191744654781,
"learning_rate": 8.322211276536733e-06,
"loss": 0.1948,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.3152721870783948,
"learning_rate": 8.27634534882907e-06,
"loss": 0.2108,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.28728912437468745,
"learning_rate": 8.230573163151585e-06,
"loss": 0.223,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.31474214454929106,
"learning_rate": 8.184895085497466e-06,
"loss": 0.2025,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.3043302678580488,
"learning_rate": 8.139311481107401e-06,
"loss": 0.2099,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.2884020103275316,
"learning_rate": 8.093822714466686e-06,
"loss": 0.211,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.2945419997171345,
"learning_rate": 8.048429149302271e-06,
"loss": 0.2216,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.29710275368705996,
"learning_rate": 8.003131148579902e-06,
"loss": 0.2183,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.28787958778598544,
"learning_rate": 7.957929074501182e-06,
"loss": 0.2027,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.29104213638558074,
"learning_rate": 7.912823288500695e-06,
"loss": 0.2175,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.3122212798900859,
"learning_rate": 7.867814151243102e-06,
"loss": 0.2152,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.3039331648398761,
"learning_rate": 7.822902022620267e-06,
"loss": 0.2034,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.2888981697484537,
"learning_rate": 7.778087261748377e-06,
"loss": 0.2124,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.3022991957078966,
"learning_rate": 7.733370226965056e-06,
"loss": 0.185,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.29447979451811884,
"learning_rate": 7.688751275826527e-06,
"loss": 0.2046,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.304576407456482,
"learning_rate": 7.644230765104735e-06,
"loss": 0.1974,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.31069947194392544,
"learning_rate": 7.599809050784502e-06,
"loss": 0.1945,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.29392540291727703,
"learning_rate": 7.555486488060661e-06,
"loss": 0.2215,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.30594710318188717,
"learning_rate": 7.511263431335249e-06,
"loss": 0.2128,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.30038846712310074,
"learning_rate": 7.467140234214656e-06,
"loss": 0.2139,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.29452838822114547,
"learning_rate": 7.423117249506777e-06,
"loss": 0.2147,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.2977002103180417,
"learning_rate": 7.379194829218231e-06,
"loss": 0.2118,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.2814975157061232,
"learning_rate": 7.335373324551518e-06,
"loss": 0.1872,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.29927044916690626,
"learning_rate": 7.291653085902224e-06,
"loss": 0.2172,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.28670248652386604,
"learning_rate": 7.248034462856213e-06,
"loss": 0.2076,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.2798679481283068,
"learning_rate": 7.20451780418683e-06,
"loss": 0.2253,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.3021576755509356,
"learning_rate": 7.161103457852128e-06,
"loss": 0.2121,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.29500681642269533,
"learning_rate": 7.117791770992046e-06,
"loss": 0.2029,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.3062176402601466,
"learning_rate": 7.074583089925688e-06,
"loss": 0.2059,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.2985860171246234,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.207,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.304593612590178,
"learning_rate": 6.988476126329604e-06,
"loss": 0.204,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.28284077681073067,
"learning_rate": 6.945578532308857e-06,
"loss": 0.2054,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.2927639648111296,
"learning_rate": 6.902785321094301e-06,
"loss": 0.2071,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.30383016030827864,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.1976,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.2801170087960711,
"learning_rate": 6.817513414939892e-06,
"loss": 0.1949,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.2933041789838248,
"learning_rate": 6.775035401831924e-06,
"loss": 0.1962,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.2902890728583524,
"learning_rate": 6.732663135188473e-06,
"loss": 0.1827,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.29269041312178684,
"learning_rate": 6.690396953817055e-06,
"loss": 0.2341,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.2845061986762335,
"learning_rate": 6.648237195676927e-06,
"loss": 0.2058,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.30903222210933573,
"learning_rate": 6.606184197876395e-06,
"loss": 0.201,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.30135558331581763,
"learning_rate": 6.564238296670109e-06,
"loss": 0.1808,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.3014562112593721,
"learning_rate": 6.52239982745638e-06,
"loss": 0.2082,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.30157225934283394,
"learning_rate": 6.480669124774484e-06,
"loss": 0.211,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.2966225531730926,
"learning_rate": 6.439046522302013e-06,
"loss": 0.2005,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.27573090654174087,
"learning_rate": 6.397532352852191e-06,
"loss": 0.219,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.2953788383959833,
"learning_rate": 6.356126948371202e-06,
"loss": 0.2152,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.2867492908981618,
"learning_rate": 6.314830639935559e-06,
"loss": 0.217,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.3043932406209416,
"learning_rate": 6.273643757749448e-06,
"loss": 0.217,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.331308485628396,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.2042,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.2916844914692928,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.2134,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.2911089056060782,
"learning_rate": 6.150742957589784e-06,
"loss": 0.1983,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.3032391564660208,
"learning_rate": 6.109997064904774e-06,
"loss": 0.2048,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.30470530717084315,
"learning_rate": 6.069362236313092e-06,
"loss": 0.2062,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.28489454486136145,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.2043,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.29075414108451053,
"learning_rate": 5.988427070179108e-06,
"loss": 0.1897,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.29066124304229235,
"learning_rate": 5.948127379792234e-06,
"loss": 0.1996,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.28940666038129864,
"learning_rate": 5.90794004780439e-06,
"loss": 0.217,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.2839232584759485,
"learning_rate": 5.867865395552405e-06,
"loss": 0.2099,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.2800181766833139,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.1841,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.30062494655418703,
"learning_rate": 5.788055411095892e-06,
"loss": 0.2293,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.28777021016925286,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.1961,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.2822714287968242,
"learning_rate": 5.708699979051613e-06,
"loss": 0.2098,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.2835727506798464,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1996,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.2932527558851661,
"learning_rate": 5.629801637510188e-06,
"loss": 0.2131,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.28648167785671536,
"learning_rate": 5.590524664836043e-06,
"loss": 0.2099,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.2790590775934915,
"learning_rate": 5.551362909942708e-06,
"loss": 0.2225,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.28070804847221825,
"learning_rate": 5.51231668596653e-06,
"loss": 0.2065,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.2842575307738721,
"learning_rate": 5.47338630512007e-06,
"loss": 0.2145,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.28040728165016926,
"learning_rate": 5.434572078689617e-06,
"loss": 0.2067,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.2988878211752244,
"learning_rate": 5.395874317032694e-06,
"loss": 0.1932,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.3005506585964586,
"learning_rate": 5.357293329575573e-06,
"loss": 0.2067,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.2908925774330441,
"learning_rate": 5.318829424810797e-06,
"loss": 0.213,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.27444297407133805,
"learning_rate": 5.280482910294728e-06,
"loss": 0.2088,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.30223563489690713,
"learning_rate": 5.242254092645078e-06,
"loss": 0.1935,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.3029176084924397,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.1926,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.3017484319457384,
"learning_rate": 5.166150769707892e-06,
"loss": 0.1866,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.2795949747823861,
"learning_rate": 5.128276872940503e-06,
"loss": 0.223,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.27080024328766034,
"learning_rate": 5.090521890074953e-06,
"loss": 0.2319,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.28408155172247634,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.2192,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.29106697155802314,
"learning_rate": 5.015369872647442e-06,
"loss": 0.1877,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.27903148869020833,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.2111,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.2977636673662977,
"learning_rate": 4.940697121074675e-06,
"loss": 0.2146,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.2866380411568811,
"learning_rate": 4.903541216934864e-06,
"loss": 0.1983,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.29312690107494604,
"learning_rate": 4.8665060236772e-06,
"loss": 0.2202,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.27932752610302125,
"learning_rate": 4.829591837434102e-06,
"loss": 0.2263,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.29400888764903355,
"learning_rate": 4.792798953370424e-06,
"loss": 0.2119,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.2821601895385218,
"learning_rate": 4.756127665681083e-06,
"loss": 0.211,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.28759986621663947,
"learning_rate": 4.719578267588725e-06,
"loss": 0.192,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.28253778343163494,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.2206,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.4548782645455139,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1484,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.38193193111220625,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1455,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.29772063380197783,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1504,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.2770541209900755,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.1342,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.35118278802674346,
"learning_rate": 4.502857857942204e-06,
"loss": 0.1459,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.4278418239593426,
"learning_rate": 4.467169815100549e-06,
"loss": 0.141,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.4505483432240444,
"learning_rate": 4.431605972355137e-06,
"loss": 0.1493,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.37179264810354695,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1494,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.3045321125101524,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1418,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.3842354500510545,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1756,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.3940388747427257,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1588,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.3429765322071072,
"learning_rate": 4.255659678050063e-06,
"loss": 0.1403,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.3133163632774194,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1456,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.27889120033813747,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1516,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.27151839931038,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1285,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.3607668737165863,
"learning_rate": 4.11716698310493e-06,
"loss": 0.1518,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.3297536319117456,
"learning_rate": 4.082860616555848e-06,
"loss": 0.1475,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.32364342988714784,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1507,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.2812217152312118,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1459,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.27226343363740846,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1719,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.2781601517409522,
"learning_rate": 3.94691060843287e-06,
"loss": 0.1519,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.279390798309327,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1666,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.31019269518084136,
"learning_rate": 3.879704680319503e-06,
"loss": 0.159,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.2824989451801182,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1338,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.27108561096701317,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1599,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.27570912224454547,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1454,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.26347546083052353,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.1467,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.2966731008883086,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.138,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.2593889511695706,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1459,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.27316207870265724,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1613,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.26337482019547637,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1599,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.2753902102971875,
"learning_rate": 3.583689060687079e-06,
"loss": 0.1753,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.2585395903370462,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.1492,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.2627997919464417,
"learning_rate": 3.51934502719796e-06,
"loss": 0.1471,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.2618970867219665,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.1557,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.25861410279761543,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1545,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.2626176439834364,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.1492,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.2715293260502499,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.1463,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.25261786914531487,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.142,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.26266988359133747,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1481,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.265476782548918,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.1443,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.27121044296288965,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.1445,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.2711609402574102,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1648,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.26482104365435616,
"learning_rate": 3.205572144561604e-06,
"loss": 0.1672,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.2537219231948773,
"learning_rate": 3.174929343376374e-06,
"loss": 0.1428,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.25795721077776496,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1427,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.261192447840744,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.1508,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.25962900311138054,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.1785,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.2646353855591152,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.1302,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.26254311677466796,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1616,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.25867865653309047,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.1506,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.2570404709180773,
"learning_rate": 2.964210239224696e-06,
"loss": 0.1558,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.26577684052607325,
"learning_rate": 2.934650470499869e-06,
"loss": 0.1577,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.24740449318488067,
"learning_rate": 2.905227155855128e-06,
"loss": 0.124,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.2504790252815796,
"learning_rate": 2.875940530558519e-06,
"loss": 0.1567,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.2593594968804313,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.1498,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.26946247889974306,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.1542,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.2613038206432522,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1707,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.25283534319165735,
"learning_rate": 2.760165589921184e-06,
"loss": 0.1295,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.2590204373539577,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.1345,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.26007559516310214,
"learning_rate": 2.703104292151164e-06,
"loss": 0.1614,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.273953461007008,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.1558,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.2557195475244978,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.1608,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.2504662861852862,
"learning_rate": 2.618550200289303e-06,
"loss": 0.1424,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.24323021639994205,
"learning_rate": 2.59064316684412e-06,
"loss": 0.1483,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.2510668372606647,
"learning_rate": 2.562875338152264e-06,
"loss": 0.151,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.24765323457284402,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1418,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.25589788221676374,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.1517,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.2584090219446137,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1458,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.2626126377341111,
"learning_rate": 2.45320049480108e-06,
"loss": 0.16,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.25747688192033086,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.1527,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.25073734381127327,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1712,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.25732066984582375,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.1389,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.26406726043128964,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.1462,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.25814578280359135,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1448,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.2657048268874292,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.1485,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.2584009298832797,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1439,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.2594739796154439,
"learning_rate": 2.24059906434998e-06,
"loss": 0.1325,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.25415518824489186,
"learning_rate": 2.214660694699091e-06,
"loss": 0.1339,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.25795320702369806,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1423,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.2530198031497636,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1482,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.25987633950107447,
"learning_rate": 2.137699676264171e-06,
"loss": 0.1373,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2561462615167252,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.1618,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.2571753713339788,
"learning_rate": 2.08710612114857e-06,
"loss": 0.152,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.2520439678763602,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.138,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.24838239756235408,
"learning_rate": 2.0370854889762e-06,
"loss": 0.1386,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.24991871307903496,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.149,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.238342470678491,
"learning_rate": 1.987639379598427e-06,
"loss": 0.1312,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.25779126666531904,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1498,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.254732086392767,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.142,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.24408091728074977,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.154,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.24595658164621326,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.1602,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.24821385584671496,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.1379,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.2686975219221582,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1647,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.267149411227535,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.1564,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.2583407324771293,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.1606,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.2507984444109105,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.1617,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.23714735766431744,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.1686,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.25172186013192266,
"learning_rate": 1.726025117089718e-06,
"loss": 0.1389,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.24493197819577414,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1201,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.25617061202272395,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1521,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.24552145512950455,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1446,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.26405639213275245,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.141,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.23378294032207503,
"learning_rate": 1.612939220355272e-06,
"loss": 0.1582,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.26543621967147785,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1489,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.26004529263557435,
"learning_rate": 1.568732761074121e-06,
"loss": 0.1384,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.2450519562149594,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1335,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.24287035290473272,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.1615,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.2563118509039775,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.1489,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.2590360903554493,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.1533,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.24853119946287067,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1443,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.2577565111258479,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.1637,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.25005955646455974,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1767,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.25553344484086465,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.1407,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.26121242543572276,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.1347,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.2423665369870452,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.137,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.2505957803965369,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1418,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.25759475801840603,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.1684,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.24519559021698284,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.1385,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.25696242195835567,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1545,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.267082321604322,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1508,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.2547010432322351,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1359,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.25113781611346014,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1578,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.2504252230313185,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1437,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.23161913726363492,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.1563,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.2496989140635014,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.1459,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.2450769782239061,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1333,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.25752154600570404,
"learning_rate": 1.121666393356604e-06,
"loss": 0.1495,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.25893371555155625,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.1412,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.24691696755680204,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.1459,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.2705369909143541,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1468,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.2553985753630126,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1537,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.2520064199803321,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1634,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2463007117519758,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1439,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.24675962215186104,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1476,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.25613579628139216,
"learning_rate": 9.771228098856534e-07,
"loss": 0.1679,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.25864695237437285,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1547,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.2575555405598959,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1588,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.32495101951478306,
"learning_rate": 9.254248023435864e-07,
"loss": 0.1747,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.25355878789715774,
"learning_rate": 9.084969902482288e-07,
"loss": 0.15,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.24677947523273938,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1573,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.24553189750071192,
"learning_rate": 8.750994656392109e-07,
"loss": 0.1578,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.24906850425319854,
"learning_rate": 8.58630020171265e-07,
"loss": 0.1615,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.25223915166954297,
"learning_rate": 8.423136285876099e-07,
"loss": 0.1734,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.249518065327455,
"learning_rate": 8.26150421353682e-07,
"loss": 0.1458,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.2566770800509792,
"learning_rate": 8.101405277100549e-07,
"loss": 0.1339,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.26174559210995346,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1466,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.2438867767741971,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1301,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.26499251020162823,
"learning_rate": 7.630320023322024e-07,
"loss": 0.1258,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.25515397798966927,
"learning_rate": 7.476366309223903e-07,
"loss": 0.1469,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.25056535360394994,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1496,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.24588216571294091,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1359,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.26710145528135837,
"learning_rate": 7.023746512481989e-07,
"loss": 0.1592,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.2537431136368797,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1476,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.2506410608903194,
"learning_rate": 6.729713135577687e-07,
"loss": 0.1599,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.2516455498107355,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1421,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.2487657045206476,
"learning_rate": 6.441861282571115e-07,
"loss": 0.1533,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.24031882746047603,
"learning_rate": 6.300256312324471e-07,
"loss": 0.1426,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.254337863670816,
"learning_rate": 6.160200160066776e-07,
"loss": 0.1575,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.24923259779801482,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1447,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.2556096246336749,
"learning_rate": 5.884738776666044e-07,
"loss": 0.1445,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.25513123488121797,
"learning_rate": 5.749335748104878e-07,
"loss": 0.1277,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.3185922463976874,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1393,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.25611353303350426,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1449,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.23625980054406967,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1274,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.254298743174193,
"learning_rate": 5.22326650565701e-07,
"loss": 0.1354,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.24672198826319652,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1517,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.24992672223481247,
"learning_rate": 4.969572286496594e-07,
"loss": 0.1351,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.25847885475072824,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1305,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.2515734506281877,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1256,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.25609021029606205,
"learning_rate": 4.600729351261857e-07,
"loss": 0.1525,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.24948151721844342,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.1434,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.2394850944965823,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1452,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.25643320858240115,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1322,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.25105966558378034,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1485,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.2515154957786091,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1526,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.25796865024204674,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1444,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.2540378123079243,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.1609,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.2746959097496441,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1489,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.25250317170659964,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1367,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.26220223824463296,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1447,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.2462418770663936,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1758,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.27737903829772775,
"learning_rate": 3.266223765343224e-07,
"loss": 0.155,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.2508843723474712,
"learning_rate": 3.16521967718757e-07,
"loss": 0.1465,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.2523734550477756,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.1433,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.24379431618429134,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1547,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.2613258574432973,
"learning_rate": 2.871653894244597e-07,
"loss": 0.16,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.25447880762858216,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.1297,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.24371305680126923,
"learning_rate": 2.683823233205485e-07,
"loss": 0.1648,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.2585593130474692,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1571,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.25399162802841235,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1526,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.2626031272908115,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1453,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.24723771221686336,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.1525,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.25401393891123086,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.141,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.2534918096605436,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1405,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.253038829539254,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1465,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.2511962610568541,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.1469,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.2598042562529633,
"learning_rate": 1.916764975761254e-07,
"loss": 0.1582,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.25975932307369975,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1573,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.2497437702667398,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.145,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.25100415723979264,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.1525,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2540832842796248,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.159,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.23933405933221558,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.1428,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.246944653708252,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.1481,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.25832608319293404,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.1353,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.24361570442189212,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1632,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.2435883184348151,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1638,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.25545953999815146,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1384,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.27061337076192177,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1527,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.2519427817083055,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1275,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.24118196471678077,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.1493,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2434253460847674,
"learning_rate": 9.787081204466253e-08,
"loss": 0.173,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.2463317454207647,
"learning_rate": 9.236233586200627e-08,
"loss": 0.1438,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2494287018087919,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1531,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.2578541133105995,
"learning_rate": 8.182296909035047e-08,
"loss": 0.1494,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.23979883770933974,
"learning_rate": 7.679216277384439e-08,
"loss": 0.1474,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.24914307578545236,
"learning_rate": 7.19206618963053e-08,
"loss": 0.1565,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.24283781314833042,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1541,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.2541578834378793,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1329,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.26474427277751483,
"learning_rate": 5.826237505053067e-08,
"loss": 0.1392,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.2623417453724291,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1555,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.26356488264217187,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1597,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.23582534081388473,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1425,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.24005530036468467,
"learning_rate": 4.228381947004812e-08,
"loss": 0.1313,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.2472450211000796,
"learning_rate": 3.868805732811343e-08,
"loss": 0.1625,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.25402821454317986,
"learning_rate": 3.525190530455636e-08,
"loss": 0.141,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.25845880873297333,
"learning_rate": 3.197539087475887e-08,
"loss": 0.1313,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.26346741574281524,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.1586,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.2463496271329015,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.1559,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.25901096522766076,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.1511,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.2382530195778275,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.1498,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.25174548986010503,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1603,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.24740947854242706,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.1436,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.24103279879811984,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1469,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.24954636671737412,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1497,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.2323845409050378,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1638,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.2602133142253077,
"learning_rate": 7.995445897566178e-09,
"loss": 0.132,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.25400612787686444,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1407,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.24638296150196648,
"learning_rate": 5.11720812557348e-09,
"loss": 0.1277,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.24372117573517074,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1402,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.2531981720563389,
"learning_rate": 2.878483275170485e-09,
"loss": 0.1472,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.2562595366219444,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1589,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.2534624507951359,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1527,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.23929942417194663,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1472,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.2482147416487185,
"learning_rate": 3.198382947511292e-10,
"loss": 0.1466,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.26351319849659377,
"learning_rate": 7.995973352770137e-11,
"loss": 0.1519,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.251476200250066,
"learning_rate": 0.0,
"loss": 0.1461,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1.3242810550997484e+18,
"train_loss": 0.32590865836090405,
"train_runtime": 15388.9045,
"train_samples_per_second": 10.267,
"train_steps_per_second": 0.08
}
],
"logging_steps": 1.0,
"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": 1.3242810550997484e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}