b1_code_top_4 / trainer_state.json
neginr's picture
End of training
0a0e0e0 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": 6.631211025687921,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.065,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.792240019504043,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0677,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.720160797668796,
"learning_rate": 9.67741935483871e-07,
"loss": 1.0613,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.6151393310292255,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.0636,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.324140464968663,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.0649,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 6.276265604919285,
"learning_rate": 1.935483870967742e-06,
"loss": 1.0593,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 4.952939235409749,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.02,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.648984777717597,
"learning_rate": 2.580645161290323e-06,
"loss": 1.0104,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.887186898813131,
"learning_rate": 2.903225806451613e-06,
"loss": 0.964,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.730728640833047,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9392,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.614023469291542,
"learning_rate": 3.548387096774194e-06,
"loss": 0.964,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.777820186081932,
"learning_rate": 3.870967741935484e-06,
"loss": 0.9162,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.0048104309920936,
"learning_rate": 4.193548387096774e-06,
"loss": 0.9328,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 3.7553038582576024,
"learning_rate": 4.516129032258065e-06,
"loss": 0.9272,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.5651572563389204,
"learning_rate": 4.838709677419355e-06,
"loss": 0.8909,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.5870522040832293,
"learning_rate": 5.161290322580646e-06,
"loss": 0.8579,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.426500912846801,
"learning_rate": 5.483870967741935e-06,
"loss": 0.8346,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.0998470910107896,
"learning_rate": 5.806451612903226e-06,
"loss": 0.8515,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.5695891491378244,
"learning_rate": 6.129032258064517e-06,
"loss": 0.8272,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.3257761110044237,
"learning_rate": 6.451612903225806e-06,
"loss": 0.8002,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.1805014259411828,
"learning_rate": 6.774193548387097e-06,
"loss": 0.7968,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.2919222813720184,
"learning_rate": 7.096774193548388e-06,
"loss": 0.778,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.1884262758250108,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.7704,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.0134850852379687,
"learning_rate": 7.741935483870968e-06,
"loss": 0.7649,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.8604164538204894,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7734,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.9908753805435532,
"learning_rate": 8.387096774193549e-06,
"loss": 0.752,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.8954841408519759,
"learning_rate": 8.70967741935484e-06,
"loss": 0.7119,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.8800222335426093,
"learning_rate": 9.03225806451613e-06,
"loss": 0.7416,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.7542277767238744,
"learning_rate": 9.35483870967742e-06,
"loss": 0.7214,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.8131137255363738,
"learning_rate": 9.67741935483871e-06,
"loss": 0.7096,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.9035308567271112,
"learning_rate": 1e-05,
"loss": 0.7236,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.7852108766438949,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.7228,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.6926036910148877,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.7225,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.6957043597859691,
"learning_rate": 1.096774193548387e-05,
"loss": 0.7036,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.6223654022493788,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.7028,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.659344923475061,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.7056,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.690557175468813,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.7144,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6565921707622225,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.7136,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.6542751372400815,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.701,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.5847384529246228,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6932,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.7022353139193868,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6632,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.555107309310691,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.6866,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.7050755609106727,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.6793,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 1.2532300577845983,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.7202,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.6374371159584226,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6825,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.6602654405200862,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6785,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.5416709413396326,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6726,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.5759134548914159,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6807,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.5434245825178284,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.6574,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.6916497701868978,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6758,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.6171878878870729,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6574,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.559757506311982,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6778,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.7183477331743063,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.677,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.5396725238485549,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6705,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.846420124029686,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6632,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.5795161241736675,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6696,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.7775124320649661,
"learning_rate": 1.838709677419355e-05,
"loss": 0.656,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.5841166463345907,
"learning_rate": 1.870967741935484e-05,
"loss": 0.663,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.5518920461253138,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6575,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.6094177067609955,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6683,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.5787170838295325,
"learning_rate": 1.967741935483871e-05,
"loss": 0.6616,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.6621180554972853,
"learning_rate": 2e-05,
"loss": 0.6604,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.5540656026092816,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6543,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.5630721763235796,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6587,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.7177719654718926,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.6472,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.6362498380223882,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6456,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.6707250119585174,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6628,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.7191816815348362,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6437,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.7961726693181498,
"learning_rate": 2.225806451612903e-05,
"loss": 0.6569,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.8533772340100138,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6566,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.8340054433589417,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6538,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.1913895949577566,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6514,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.8109161056523392,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6275,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.1073587375421188,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6485,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.003345595733349,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6516,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.9102247522992614,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.6449,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.2265853521935515,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6558,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.205945109107846,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.641,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.9306470874459888,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6481,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.703401547056361,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6532,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.267411079982747,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.6256,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.6448957377659374,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.6498,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.4923213490572527,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6587,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.306630800714686,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.6524,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 4.214443283082755,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.65,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.3636439062116656,
"learning_rate": 2.774193548387097e-05,
"loss": 0.6442,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.2845447306373778,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.6457,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.9516229041705945,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.6352,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.8794157664691558,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.6453,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.122721064449745,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.641,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.0269721045071614,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6503,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 135.73849738706767,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6603,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.5375798753990426,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.6504,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.0608533798435056,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6459,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.2193735301089295,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.6322,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.2249522827002952,
"learning_rate": 3.096774193548387e-05,
"loss": 0.6319,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.0955312040734968,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.6145,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.0869403709516272,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6476,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.7779966916229196,
"learning_rate": 3.193548387096774e-05,
"loss": 0.6381,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.1170464037858976,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.6313,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.1102090961687774,
"learning_rate": 3.258064516129033e-05,
"loss": 0.6409,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.6865269901949911,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6294,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.1698676656096054,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6444,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.8718432675365414,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.6441,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.881111920192446,
"learning_rate": 3.387096774193549e-05,
"loss": 0.6406,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.9779128150516884,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.6408,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.9366462009517272,
"learning_rate": 3.451612903225807e-05,
"loss": 0.6421,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.3463698019635424,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6304,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.9210354360517241,
"learning_rate": 3.516129032258065e-05,
"loss": 0.6447,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.6882555305352531,
"learning_rate": 3.548387096774194e-05,
"loss": 0.6324,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.3218382956589656,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.6378,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.2548146444721657,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6335,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.2301173493131474,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.6527,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.0654799430670485,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6308,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.2166613558388544,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6337,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.0658529654308224,
"learning_rate": 3.741935483870968e-05,
"loss": 0.6337,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 1.0168735995685234,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.6384,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.2791686765834325,
"learning_rate": 3.806451612903226e-05,
"loss": 0.6313,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.5283134971227006,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.6466,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.943229120232067,
"learning_rate": 3.870967741935484e-05,
"loss": 0.6311,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.780402139280714,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.6499,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.0333931459248058,
"learning_rate": 3.935483870967742e-05,
"loss": 0.6327,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.5537205322676695,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.6274,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.1641450390261672,
"learning_rate": 4e-05,
"loss": 0.6452,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.3614833177345855,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.631,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.1213708413540648,
"learning_rate": 3.999968016170525e-05,
"loss": 0.6361,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.1030438881168587,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.6406,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.4483465837279683,
"learning_rate": 3.999872065705065e-05,
"loss": 0.63,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.9044141350716015,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6328,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.3342442886988366,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6359,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.4469329399650013,
"learning_rate": 3.999608209836944e-05,
"loss": 0.6265,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.0751947789349312,
"learning_rate": 3.999488279187443e-05,
"loss": 0.6337,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.7111452013650859,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6336,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.3986665841614252,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.6445,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.3422667232204286,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6441,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.0590563320777693,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6246,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 1.1299909376511759,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6231,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.8942888169300758,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6195,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 1.149809906107642,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6292,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.7498186090888078,
"learning_rate": 3.997953378607961e-05,
"loss": 0.6363,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.8760490641830024,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6246,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.9619021247969958,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6302,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.8075791156695534,
"learning_rate": 3.997114145976237e-05,
"loss": 0.6196,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.7973432565863953,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.6278,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.6769867038046066,
"learning_rate": 3.996474809469545e-05,
"loss": 0.621,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.7202813820102747,
"learning_rate": 3.996131194267189e-05,
"loss": 0.6267,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.7975103614222211,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6141,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.6320619304757701,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6302,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.8825536925793266,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6107,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.7469654076500898,
"learning_rate": 3.994597152728987e-05,
"loss": 0.6135,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.7552338674995533,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6208,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.7682039203816183,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.6066,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.8425961126560738,
"learning_rate": 3.993279149458988e-05,
"loss": 0.628,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.859284400590846,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.6157,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.6106342232876313,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6393,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.852949860848789,
"learning_rate": 3.991817703090965e-05,
"loss": 0.6313,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.6306314566837432,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6198,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.6676648575256504,
"learning_rate": 3.9907637664138e-05,
"loss": 0.6186,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.5816533265343266,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6175,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.773758898340095,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.6156,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.631109253065057,
"learning_rate": 3.989063487022416e-05,
"loss": 0.6346,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.911838474102337,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.6309,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.6859925835827286,
"learning_rate": 3.987850437381915e-05,
"loss": 0.6253,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.8679336029473558,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6305,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.7297947274115318,
"learning_rate": 3.986573808672002e-05,
"loss": 0.6301,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.678648200395756,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.6254,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.7751562160932809,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6123,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.7626307788597796,
"learning_rate": 3.984539744726554e-05,
"loss": 0.629,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.6886135665994733,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6182,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.8991421728840339,
"learning_rate": 3.983104351425871e-05,
"loss": 0.628,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.5339363303205484,
"learning_rate": 3.982362866600154e-05,
"loss": 0.6248,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.8084237241172356,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.6273,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.8643198594759519,
"learning_rate": 3.980832350242388e-05,
"loss": 0.6036,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.5143001434391795,
"learning_rate": 3.980043330948306e-05,
"loss": 0.6226,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.8440199408863964,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6176,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.681982195879739,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6199,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.6385713218354864,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.6197,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.8394280762241233,
"learning_rate": 3.976728995023254e-05,
"loss": 0.6327,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.6921306610346109,
"learning_rate": 3.975860879481514e-05,
"loss": 0.6265,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.9111509109386063,
"learning_rate": 3.974976965008833e-05,
"loss": 0.6201,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.7498404817972449,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6297,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.613952154493775,
"learning_rate": 3.973161767667946e-05,
"loss": 0.622,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.8105001214355401,
"learning_rate": 3.972230499314009e-05,
"loss": 0.5943,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.6933868849420868,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.6341,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.5765832420247458,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.628,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.6154461732185325,
"learning_rate": 3.969342105253098e-05,
"loss": 0.6162,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.6431050381255247,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.6353,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 1.3361008003423989,
"learning_rate": 3.967337762346568e-05,
"loss": 0.6375,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 2.032370644345463,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.6163,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.7069465860237731,
"learning_rate": 3.965270496444528e-05,
"loss": 0.6368,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.5654826725644035,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.6298,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.6137944001369037,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.6179,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.5619535620038694,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6183,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.5219087168253574,
"learning_rate": 3.960947462140638e-05,
"loss": 0.6181,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.5890267631717163,
"learning_rate": 3.959827482437472e-05,
"loss": 0.6193,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.5650316749178916,
"learning_rate": 3.958691832005981e-05,
"loss": 0.6169,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.6677079640616458,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.6234,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.5464017627899661,
"learning_rate": 3.956373555405774e-05,
"loss": 0.6161,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.6056708138539372,
"learning_rate": 3.955190947773937e-05,
"loss": 0.6294,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.640233636940533,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.6014,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.5431917464121101,
"learning_rate": 3.952778841127214e-05,
"loss": 0.6142,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.6286107745331339,
"learning_rate": 3.95154936139947e-05,
"loss": 0.6087,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.5496904064096922,
"learning_rate": 3.950304277135034e-05,
"loss": 0.6051,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.5854422864591703,
"learning_rate": 3.94904359828957e-05,
"loss": 0.6112,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.5030569772170853,
"learning_rate": 3.94776733494343e-05,
"loss": 0.6182,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.5018853854841132,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.6012,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.48056943804150215,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.6067,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.5680064694557975,
"learning_rate": 3.943845140573213e-05,
"loss": 0.6091,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.42435528519041216,
"learning_rate": 3.942506642518952e-05,
"loss": 0.6259,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.5827809714572664,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.6104,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.49154196545067724,
"learning_rate": 3.939783060543167e-05,
"loss": 0.5964,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5341777658149737,
"learning_rate": 3.938397998399333e-05,
"loss": 0.6076,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.4922675439319432,
"learning_rate": 3.936997436876756e-05,
"loss": 0.6186,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.5209908638146832,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5996,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.5532204049668288,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.6087,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.5666782596523825,
"learning_rate": 3.932702868644224e-05,
"loss": 0.6056,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.5043936602124267,
"learning_rate": 3.931240422833181e-05,
"loss": 0.6176,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.6425050963904512,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.6025,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.49059309501842435,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.6047,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6060984582161967,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.6065,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.6285164673691697,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.6016,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.60568101340624,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.6134,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.5246974808745659,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.6033,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.5967537308799818,
"learning_rate": 3.92057159243286e-05,
"loss": 0.6126,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.5721163921846537,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.5913,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.6258604885103458,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.6179,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.5229877210547087,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.6193,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.5748458979341448,
"learning_rate": 3.914136997982874e-05,
"loss": 0.6102,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.5630062119398437,
"learning_rate": 3.912490053436079e-05,
"loss": 0.599,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.5205480530263807,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5915,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.5822410746485841,
"learning_rate": 3.909150300975178e-05,
"loss": 0.6076,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.5568522737326885,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5886,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.6053057161264237,
"learning_rate": 3.905749486576607e-05,
"loss": 0.5844,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.5360951133439625,
"learning_rate": 3.90402621506546e-05,
"loss": 0.5961,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.6089760914242993,
"learning_rate": 3.902287719011435e-05,
"loss": 0.6131,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.6067188523431459,
"learning_rate": 3.900534012315499e-05,
"loss": 0.6052,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5441380021255628,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.6237,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.6173625770878225,
"learning_rate": 3.896981023209777e-05,
"loss": 0.6067,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.5810327234292192,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.6139,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.6781484848205109,
"learning_rate": 3.893367361386491e-05,
"loss": 0.6074,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.5050464875981803,
"learning_rate": 3.891537814248417e-05,
"loss": 0.6074,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.6020784654050509,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5951,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.5593208501740763,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.6005,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.5475170198835017,
"learning_rate": 3.885958483839049e-05,
"loss": 0.5986,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.5264777726192841,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.6008,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.6300316910505718,
"learning_rate": 3.882163505079171e-05,
"loss": 0.6106,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.5973340925263346,
"learning_rate": 3.880243433489132e-05,
"loss": 0.6112,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 4.023710689702117,
"learning_rate": 3.878308327522702e-05,
"loss": 0.5857,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.8336474545983982,
"learning_rate": 3.876358202652936e-05,
"loss": 0.5366,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.7274857827151454,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.5451,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.8456075150808504,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.5423,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.7062502267893811,
"learning_rate": 3.870417871154801e-05,
"loss": 0.5273,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.7000124618442125,
"learning_rate": 3.868407827802192e-05,
"loss": 0.5294,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.8078669979033686,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.518,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.611257761811159,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.52,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.7824732320012349,
"learning_rate": 3.862288124195319e-05,
"loss": 0.5485,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.5429900635293294,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.5529,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.7146825803527466,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.5336,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.48182140593526285,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.5463,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.6708417528093994,
"learning_rate": 3.853920126717703e-05,
"loss": 0.5232,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.5850174906555812,
"learning_rate": 3.851791025492574e-05,
"loss": 0.5552,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.6012549969223946,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.5256,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.5032941945961664,
"learning_rate": 3.847488419569866e-05,
"loss": 0.5336,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.595272481958689,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.5323,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.6569768500079539,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.5439,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.6085114385692786,
"learning_rate": 3.840923760917199e-05,
"loss": 0.53,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.5429191081618998,
"learning_rate": 3.838706077964473e-05,
"loss": 0.5215,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.6321146053177994,
"learning_rate": 3.836473692766944e-05,
"loss": 0.545,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.6941528028510345,
"learning_rate": 3.834226623174705e-05,
"loss": 0.5291,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.4724793767254631,
"learning_rate": 3.831964887155264e-05,
"loss": 0.5322,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.563551877506454,
"learning_rate": 3.829688502793403e-05,
"loss": 0.5092,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.6715096754619526,
"learning_rate": 3.827397488291029e-05,
"loss": 0.5407,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.5640667342601671,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.4998,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.5795912289027252,
"learning_rate": 3.822771642257145e-05,
"loss": 0.5234,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.7051581161445596,
"learning_rate": 3.820436847713776e-05,
"loss": 0.5244,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.5661115578559703,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.528,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.4128558510245345,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.5283,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.6851380904383566,
"learning_rate": 3.813345202354145e-05,
"loss": 0.5333,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 1.0358206824831324,
"learning_rate": 3.810952296329563e-05,
"loss": 0.5527,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.4757187365049208,
"learning_rate": 3.808544909978676e-05,
"loss": 0.5397,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.7362949899764445,
"learning_rate": 3.806123062550882e-05,
"loss": 0.5363,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.5171829522722434,
"learning_rate": 3.803686773411208e-05,
"loss": 0.5333,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.6404699490367847,
"learning_rate": 3.801236062040158e-05,
"loss": 0.5497,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.685824703551068,
"learning_rate": 3.798770948033553e-05,
"loss": 0.5426,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.43455009822313906,
"learning_rate": 3.796291451102381e-05,
"loss": 0.5265,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.6428361006507286,
"learning_rate": 3.793797591072631e-05,
"loss": 0.5093,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.6554408395817922,
"learning_rate": 3.791289387885144e-05,
"loss": 0.5273,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.5072884226301756,
"learning_rate": 3.788766861595443e-05,
"loss": 0.5295,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.48400261504010356,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.542,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.5893210008692044,
"learning_rate": 3.783678920503983e-05,
"loss": 0.5406,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.45241088389997075,
"learning_rate": 3.781113546385265e-05,
"loss": 0.5446,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.5273584399424653,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.5364,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.5587103581141739,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.538,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.40595831232274343,
"learning_rate": 3.77333205623025e-05,
"loss": 0.518,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.4623259424651416,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.5229,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.44922253303052473,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.543,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.4306090352006099,
"learning_rate": 3.765422951112704e-05,
"loss": 0.5501,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.5053153463520782,
"learning_rate": 3.762758321970216e-05,
"loss": 0.5132,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.4972981324988089,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.5466,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.5289062762978823,
"learning_rate": 3.757386800198538e-05,
"loss": 0.5314,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.4926455563649697,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.5367,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.6663288752788136,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.5287,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.5713223883483739,
"learning_rate": 3.749224181796305e-05,
"loss": 0.5344,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5186861496735482,
"learning_rate": 3.746475306375545e-05,
"loss": 0.5385,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.5544509390298846,
"learning_rate": 3.743712466184774e-05,
"loss": 0.5328,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.4605247500739111,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.5396,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.4785301342956885,
"learning_rate": 3.73814497997107e-05,
"loss": 0.5387,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.4918329640597997,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.534,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.42406097077537463,
"learning_rate": 3.732521901224724e-05,
"loss": 0.5291,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.49006661976700056,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.5358,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.4631612639246261,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.5407,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.45661733226218515,
"learning_rate": 3.723983441007882e-05,
"loss": 0.5277,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.4814918014756271,
"learning_rate": 3.721109687296781e-05,
"loss": 0.5376,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.44805361775373487,
"learning_rate": 3.718222171638484e-05,
"loss": 0.5262,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.5012446111019487,
"learning_rate": 3.715320917121488e-05,
"loss": 0.5387,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.40719711510689804,
"learning_rate": 3.712405946944149e-05,
"loss": 0.5454,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.5452634736218631,
"learning_rate": 3.709477284414488e-05,
"loss": 0.5364,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.45246875343113735,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.5123,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.5509101817444658,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.5304,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.507900506085017,
"learning_rate": 3.70060937743295e-05,
"loss": 0.5402,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.527518009797197,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.5351,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.5277861801277963,
"learning_rate": 3.694629409915555e-05,
"loss": 0.5628,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.4119167937877866,
"learning_rate": 3.691619088858401e-05,
"loss": 0.531,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.5078563180402471,
"learning_rate": 3.68859524166009e-05,
"loss": 0.523,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.4287251422813006,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.5501,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.5024968192967032,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.5384,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.4746927339147861,
"learning_rate": 3.67944278554384e-05,
"loss": 0.5476,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.4384452599139998,
"learning_rate": 3.676365076645557e-05,
"loss": 0.5333,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.4422284397463881,
"learning_rate": 3.673273963576791e-05,
"loss": 0.5376,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.47173074550929583,
"learning_rate": 3.670169471054001e-05,
"loss": 0.5336,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.5191034211880015,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.5296,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.41626602228581666,
"learning_rate": 3.663920447046888e-05,
"loss": 0.5331,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.5240861874453606,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.53,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.5542420658731266,
"learning_rate": 3.657618204491937e-05,
"loss": 0.5285,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.42644212576664187,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.5096,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.47832239085341216,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.541,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.45358995547900055,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.5421,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.4228297526508118,
"learning_rate": 3.644854871713611e-05,
"loss": 0.5321,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.5756659246941286,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.537,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.4197179024308734,
"learning_rate": 3.638394189710493e-05,
"loss": 0.5367,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.47286793957725937,
"learning_rate": 3.635144184933412e-05,
"loss": 0.5296,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5068654291062417,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.5502,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.44824290418921153,
"learning_rate": 3.628604977762752e-05,
"loss": 0.5336,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.3975879958219806,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.5309,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.4544462630040234,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.5476,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.4278190779369588,
"learning_rate": 3.618698565901727e-05,
"loss": 0.554,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.42641555948966303,
"learning_rate": 3.615370507164654e-05,
"loss": 0.5245,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.3987816308184671,
"learning_rate": 3.61202953196805e-05,
"loss": 0.5503,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.5077233260650009,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.5368,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.5005367629457359,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.546,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.4758648540561926,
"learning_rate": 3.601929375279662e-05,
"loss": 0.5482,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.4415819044402995,
"learning_rate": 3.598537002418012e-05,
"loss": 0.5453,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.41070552660326864,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.5368,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.4067231316665983,
"learning_rate": 3.591713938344416e-05,
"loss": 0.5164,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.45047438892879926,
"learning_rate": 3.588283301689507e-05,
"loss": 0.5353,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.5190465762952905,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.5383,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.39555567826702676,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.5203,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.5078711394790769,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.5321,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.43296670153698025,
"learning_rate": 3.574434032194994e-05,
"loss": 0.5273,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.5046334261808618,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.5373,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5042245949418402,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.5372,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.41816413737702435,
"learning_rate": 3.563914797637216e-05,
"loss": 0.529,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.4416592417321981,
"learning_rate": 3.560383338592649e-05,
"loss": 0.5269,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.4219310417323453,
"learning_rate": 3.556839402764487e-05,
"loss": 0.5265,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.42308339618293983,
"learning_rate": 3.553283018489946e-05,
"loss": 0.5193,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.4792871196369777,
"learning_rate": 3.54971421420578e-05,
"loss": 0.5392,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.4059161272623678,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.5475,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.49376802060473374,
"learning_rate": 3.54253945985191e-05,
"loss": 0.5299,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.458058540325273,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.5391,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.4094055170087161,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.538,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.4858140095012969,
"learning_rate": 3.531684894865866e-05,
"loss": 0.56,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.49894420514926,
"learning_rate": 3.528042173241128e-05,
"loss": 0.5204,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.42925347109010353,
"learning_rate": 3.524387233431892e-05,
"loss": 0.5289,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.5035964891890792,
"learning_rate": 3.520720104662958e-05,
"loss": 0.5202,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.39806684074724435,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.5372,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.4651926910054205,
"learning_rate": 3.513349397632281e-05,
"loss": 0.5278,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.3830074743923134,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.5315,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.4080543768636597,
"learning_rate": 3.505930287892533e-05,
"loss": 0.5183,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.414885732813042,
"learning_rate": 3.502202656100099e-05,
"loss": 0.545,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.42962459485706556,
"learning_rate": 3.498463012735256e-05,
"loss": 0.5254,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.44229187647103335,
"learning_rate": 3.494711387700094e-05,
"loss": 0.53,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.47426192079486895,
"learning_rate": 3.490947810992505e-05,
"loss": 0.5468,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.3975043726291194,
"learning_rate": 3.48717231270595e-05,
"loss": 0.5224,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.444773246678999,
"learning_rate": 3.483384923029211e-05,
"loss": 0.5183,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.4689405540771949,
"learning_rate": 3.479585672246156e-05,
"loss": 0.5346,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.44264466009598163,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.5331,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.5626091140755334,
"learning_rate": 3.471951708970527e-05,
"loss": 0.5307,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.44091689093259645,
"learning_rate": 3.468117057518921e-05,
"loss": 0.528,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.5198638643347055,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.5236,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.5129798611073167,
"learning_rate": 3.460412568296731e-05,
"loss": 0.532,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.4108330836577735,
"learning_rate": 3.456542792131039e-05,
"loss": 0.5311,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.4944435205196696,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.5355,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.46075630697004,
"learning_rate": 3.448768331403348e-05,
"loss": 0.55,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.4316244886796523,
"learning_rate": 3.44486370900573e-05,
"loss": 0.5167,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.5600181835660403,
"learning_rate": 3.440947533516396e-05,
"loss": 0.5276,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.4496841368729442,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5373,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.4939906328556625,
"learning_rate": 3.43308064860925e-05,
"loss": 0.5423,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.4758725052448542,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.5243,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.47992529034945175,
"learning_rate": 3.425167928295015e-05,
"loss": 0.5376,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.42391508249836857,
"learning_rate": 3.421194458890411e-05,
"loss": 0.53,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.4322584950021095,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.5325,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.44129443723200795,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.5336,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.4381137327574919,
"learning_rate": 3.409205995219561e-05,
"loss": 0.5267,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.433391336765966,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.5358,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.43230210195385027,
"learning_rate": 3.40115729298209e-05,
"loss": 0.5492,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.5114725264018753,
"learning_rate": 3.397116120327025e-05,
"loss": 0.5335,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.4080701122402877,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.5306,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.38963459503852854,
"learning_rate": 3.389000293509524e-05,
"loss": 0.5085,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.4567137083068699,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.5312,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.46103778474444035,
"learning_rate": 3.380840041143494e-05,
"loss": 0.534,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.427691848019859,
"learning_rate": 3.376743336885792e-05,
"loss": 0.5401,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.5040874850236803,
"learning_rate": 3.372635624225055e-05,
"loss": 0.5353,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.4010090738257498,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.5436,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.5458417736732056,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.5323,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.4637561562650452,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.5282,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.4389814664304234,
"learning_rate": 3.356095347769799e-05,
"loss": 0.539,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.4919210581622846,
"learning_rate": 3.351933087522552e-05,
"loss": 0.5158,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.40405594010467333,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.5185,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.4819585397285566,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.5436,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.4371743186924878,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.5384,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.4413814108146188,
"learning_rate": 3.335176280432308e-05,
"loss": 0.5132,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.51030711972321,
"learning_rate": 3.330960304618295e-05,
"loss": 0.548,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.526939670910967,
"learning_rate": 3.326733686481153e-05,
"loss": 0.5432,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.43454916366134994,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.5272,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.5735539596824841,
"learning_rate": 3.318248658506012e-05,
"loss": 0.5382,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.4183934170036395,
"learning_rate": 3.313990316514069e-05,
"loss": 0.5306,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.45045357514832596,
"learning_rate": 3.309721467890571e-05,
"loss": 0.5291,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.4720828473967064,
"learning_rate": 3.305442146769115e-05,
"loss": 0.5281,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.3508328665637698,
"learning_rate": 3.30115238736704e-05,
"loss": 0.5431,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.4820526616620955,
"learning_rate": 3.296852223985148e-05,
"loss": 0.5146,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.42337066368450427,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.5436,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.4002266958305888,
"learning_rate": 3.288220822900796e-05,
"loss": 0.5286,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.47724155529851825,
"learning_rate": 3.283889654214788e-05,
"loss": 0.5356,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.3933468639969869,
"learning_rate": 3.279548219581317e-05,
"loss": 0.5349,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.4550002073792942,
"learning_rate": 3.275196553714379e-05,
"loss": 0.545,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.4161950026451521,
"learning_rate": 3.270834691409778e-05,
"loss": 0.5274,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.4618146046479842,
"learning_rate": 3.266462667544849e-05,
"loss": 0.5341,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.391621995595622,
"learning_rate": 3.262080517078178e-05,
"loss": 0.5238,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.47620936493638205,
"learning_rate": 3.257688275049323e-05,
"loss": 0.5264,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.4418940511754135,
"learning_rate": 3.253285976578535e-05,
"loss": 0.5097,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.4422349255333096,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.5415,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.4537886519759011,
"learning_rate": 3.244451351193935e-05,
"loss": 0.5582,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.39228114799610175,
"learning_rate": 3.240019094921551e-05,
"loss": 0.5157,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.47656552794573387,
"learning_rate": 3.235576923489527e-05,
"loss": 0.5227,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.39229544401628313,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.5177,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.4690996704963843,
"learning_rate": 3.226662977303494e-05,
"loss": 0.5434,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.36830117602801005,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.5398,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.41884804203264564,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.532,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.4117667458043893,
"learning_rate": 3.21321858487569e-05,
"loss": 0.5417,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.3553475324061862,
"learning_rate": 3.208717671149931e-05,
"loss": 0.5225,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.4011756804309766,
"learning_rate": 3.204207092549882e-05,
"loss": 0.5356,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.3844591063098084,
"learning_rate": 3.19968688514201e-05,
"loss": 0.5419,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.36562005543544823,
"learning_rate": 3.195157085069774e-05,
"loss": 0.535,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.4022684049127657,
"learning_rate": 3.190617728553332e-05,
"loss": 0.5369,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.38477208215243325,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.5265,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.41211689616116426,
"learning_rate": 3.181510491450254e-05,
"loss": 0.5351,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.4345095274648374,
"learning_rate": 3.176942683684842e-05,
"loss": 0.5262,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.4504399410592123,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.5365,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.5507463986004376,
"learning_rate": 3.167778872346327e-05,
"loss": 0.5316,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.3949407996339669,
"learning_rate": 3.163182942046815e-05,
"loss": 0.535,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.5587509189207124,
"learning_rate": 3.158577710967494e-05,
"loss": 0.5148,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.3902642479139051,
"learning_rate": 3.153963215931669e-05,
"loss": 0.5407,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.5087540026881411,
"learning_rate": 3.14933949383672e-05,
"loss": 0.5199,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.3824864662134254,
"learning_rate": 3.144706581653805e-05,
"loss": 0.5361,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.4931063160893649,
"learning_rate": 3.140064516427566e-05,
"loss": 0.5167,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.42171573533644674,
"learning_rate": 3.135413335275833e-05,
"loss": 0.5373,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.44911114830915555,
"learning_rate": 3.130753075389327e-05,
"loss": 0.5277,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.4059524470283339,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.528,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.4135398084265245,
"learning_rate": 3.121405468537547e-05,
"loss": 0.5287,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.47511626298355475,
"learning_rate": 3.116718196315488e-05,
"loss": 0.5423,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.4086318979414822,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.5166,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.5025565804010096,
"learning_rate": 3.107316901675252e-05,
"loss": 0.53,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.4429540482063417,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.5394,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.4667152706574683,
"learning_rate": 3.097880190800057e-05,
"loss": 0.532,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.3753250859068532,
"learning_rate": 3.093148648549788e-05,
"loss": 0.5405,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.4605238472487858,
"learning_rate": 3.088408365512055e-05,
"loss": 0.5189,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.46100111388408077,
"learning_rate": 3.083659379590034e-05,
"loss": 0.5364,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.5126376802556446,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.5359,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.39853276603268295,
"learning_rate": 3.074135451053475e-05,
"loss": 0.5331,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.4556659547094553,
"learning_rate": 3.069360584592015e-05,
"loss": 0.5457,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.36177466237775135,
"learning_rate": 3.064577167551816e-05,
"loss": 0.5285,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.42260143166663916,
"learning_rate": 3.059785238180953e-05,
"loss": 0.5179,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.4041162034896257,
"learning_rate": 3.054984834795566e-05,
"loss": 0.5244,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.3943909731373555,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.5382,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.4229926833743256,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.5186,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.4077656162767951,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.5515,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.42933110489325294,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.5031,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.408026347617519,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.5429,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.45274079232603176,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.5332,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3921744390640025,
"learning_rate": 3.021147971398632e-05,
"loss": 0.5076,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.47669193977416335,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.5262,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.39212924157040513,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.5426,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.5423820642116065,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.5463,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.5559238396466615,
"learning_rate": 3.001632138223332e-05,
"loss": 0.5442,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.727027487225023,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.4582,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.5583885832977757,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.4557,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.7675790966479636,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.4426,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.6786688554457859,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.4037,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.542761775118988,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.4479,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.5608485378985459,
"learning_rate": 2.972118904869376e-05,
"loss": 0.419,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.6050872219681337,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.4519,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.4962430558249029,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.4258,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.5754693571998224,
"learning_rate": 2.95725682050846e-05,
"loss": 0.417,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.5386457130742947,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.4384,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.6276593792491342,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.4578,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5672107304716811,
"learning_rate": 2.942325848714505e-05,
"loss": 0.4234,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.664799717626366,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.4638,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.5293298322211867,
"learning_rate": 2.932334127129686e-05,
"loss": 0.4297,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.6362840886455773,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.4364,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.5384007203356579,
"learning_rate": 2.922312585929131e-05,
"loss": 0.4099,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.5644327349994402,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.4527,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.4644734309990636,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.4394,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.5030023431664443,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.4241,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.46842276577279696,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.4384,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.5521576514870985,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.4261,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.4178558611275253,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.4393,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.500191881163242,
"learning_rate": 2.887006998351372e-05,
"loss": 0.4422,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.45522764909028235,
"learning_rate": 2.881934649605535e-05,
"loss": 0.4547,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.4811241923644427,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.4431,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.48535276029469765,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.4519,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.4117471740029415,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.4207,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.45596438908711423,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.3981,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.4108050406517983,
"learning_rate": 2.856467940121025e-05,
"loss": 0.4459,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.46258358073487615,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.4706,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.3932117332734876,
"learning_rate": 2.846233032790149e-05,
"loss": 0.4197,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.4035022397416761,
"learning_rate": 2.84110540896642e-05,
"loss": 0.4189,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.4253307732999496,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.4339,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.3984155086169801,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.4475,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.4523876151265892,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.4332,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.4010654124600214,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.4417,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.43109481510596015,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.4426,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.384427219365096,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.4541,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.41147388346942526,
"learning_rate": 2.805026032173773e-05,
"loss": 0.4422,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.3770796780452168,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.424,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.439842936377623,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.4536,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.4195937679853319,
"learning_rate": 2.7894660620408e-05,
"loss": 0.4082,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.4287801821251631,
"learning_rate": 2.784266724883173e-05,
"loss": 0.4332,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.47154766268852266,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.4465,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.3973185339599507,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.4442,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.5523575481210924,
"learning_rate": 2.768631254100716e-05,
"loss": 0.4581,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.42992728591404894,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.4219,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.44717409474035824,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.43,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.47353429005461234,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.4327,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.45387309932016484,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.4398,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.5304118190895447,
"learning_rate": 2.742449775134085e-05,
"loss": 0.4533,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.44607264015990944,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.4229,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.495125684268592,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.4261,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.4663930456435153,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.4181,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.4617720139636221,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.4227,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.44179296891895564,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.4612,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.47409632349073255,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.4382,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.4753204544688991,
"learning_rate": 2.705547548999996e-05,
"loss": 0.4391,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.5145808150894848,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.4279,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.42069310943674676,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.4282,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.41460930767898907,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.4365,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.4215126623676529,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.4559,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.3662047296652484,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.4438,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.3818601225263465,
"learning_rate": 2.67369650552155e-05,
"loss": 0.449,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.4305367932732089,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.4362,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.37913497544170677,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.4493,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.40057956882857276,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.4194,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.4245590212138922,
"learning_rate": 2.652354173593614e-05,
"loss": 0.4449,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.37837794048982676,
"learning_rate": 2.647005443331317e-05,
"loss": 0.4303,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.44612210888120785,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.4347,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.39139558727751156,
"learning_rate": 2.636292505301542e-05,
"loss": 0.4374,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.38852632035294804,
"learning_rate": 2.630928383194431e-05,
"loss": 0.4368,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.37984614311436654,
"learning_rate": 2.625559216200781e-05,
"loss": 0.4304,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.41196295452932985,
"learning_rate": 2.620185047252308e-05,
"loss": 0.4433,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.4242358684109573,
"learning_rate": 2.614805919320723e-05,
"loss": 0.454,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.41571814637002746,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.4232,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.3816445053050571,
"learning_rate": 2.60403295859298e-05,
"loss": 0.4348,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.4085132573929402,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.4438,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.3430607779522511,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.4162,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.38881979136479017,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.4401,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.37941756092610607,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.4395,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.361849525085354,
"learning_rate": 2.577016790136453e-05,
"loss": 0.4414,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.3984034617614385,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.4351,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.3607413093537968,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.4539,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.4133505113624862,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.4384,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.36336603722638744,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.4597,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.46413356624272006,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.444,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.381884343275129,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.4254,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.4539780242721823,
"learning_rate": 2.539001591449889e-05,
"loss": 0.4384,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.3756329837535657,
"learning_rate": 2.53355326151989e-05,
"loss": 0.409,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.4212613012784988,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.4182,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.3703218629175987,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.431,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.3587549331996114,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.4309,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.41176418102503504,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.4348,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.35983101481638774,
"learning_rate": 2.506248490185627e-05,
"loss": 0.4427,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.41171822838677513,
"learning_rate": 2.500775217241923e-05,
"loss": 0.4394,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.3608767833315591,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.4467,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.3819079915110959,
"learning_rate": 2.4898167025948e-05,
"loss": 0.4197,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.37578670486422144,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.4261,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.38164301032582604,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.4373,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.36832878246882933,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.4389,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.3503218871193688,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.4299,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.38990379925862706,
"learning_rate": 2.46235264422994e-05,
"loss": 0.4486,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.37109294389056624,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.4478,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.3692372547265715,
"learning_rate": 2.451340834511592e-05,
"loss": 0.4569,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.4004316505284065,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.4428,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.36223747704573,
"learning_rate": 2.440314589184958e-05,
"loss": 0.4282,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.39761196511060937,
"learning_rate": 2.434796163357541e-05,
"loss": 0.4375,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.3627112472118591,
"learning_rate": 2.429274260911588e-05,
"loss": 0.4349,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.39328548904469635,
"learning_rate": 2.423748926000084e-05,
"loss": 0.4347,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.3415364845697387,
"learning_rate": 2.418220202803459e-05,
"loss": 0.439,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.4128816693701262,
"learning_rate": 2.412688135529237e-05,
"loss": 0.4446,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.3713922365152482,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.4365,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.8406659493412595,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.4535,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.3654572811821014,
"learning_rate": 2.396072311715187e-05,
"loss": 0.4284,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.3968951238643986,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.4281,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.36381824452056905,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.4421,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.4099153642987632,
"learning_rate": 2.379427985200223e-05,
"loss": 0.4495,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.3692874060532849,
"learning_rate": 2.373873749394972e-05,
"loss": 0.4201,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.47110684618577287,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.4268,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.3753427882671947,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.4381,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 3.269748128247852,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.4597,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.4364777089343932,
"learning_rate": 2.351627355800233e-05,
"loss": 0.45,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.3832487886424579,
"learning_rate": 2.346058617160256e-05,
"loss": 0.4186,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.41889319058430036,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.4474,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.3988363293495803,
"learning_rate": 2.334912883203469e-05,
"loss": 0.4265,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.37827332198025193,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.4467,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.4391584398856758,
"learning_rate": 2.323756437450136e-05,
"loss": 0.44,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.3355566340050802,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.433,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.39304586189972385,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.4452,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.35252344499017363,
"learning_rate": 2.307002464849052e-05,
"loss": 0.4831,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.36032540557508624,
"learning_rate": 2.30141283818846e-05,
"loss": 0.4448,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.4065300139469985,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.4158,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.3383889886957959,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.4232,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.375422833777996,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.4396,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.33032932741861953,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.4468,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.3776060064955788,
"learning_rate": 2.273429448107257e-05,
"loss": 0.4386,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.34407629644142335,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.4212,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.351332983032895,
"learning_rate": 2.262220474347102e-05,
"loss": 0.4385,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.40274602226260997,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.4279,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.338615317050348,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.426,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.40379192973727324,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.4378,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.35514009015468817,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.4493,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.3516698976898478,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.4142,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.3603780864234384,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.4176,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.354741966333996,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.4436,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.30223582646608327,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.4648,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.3667719306768699,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.4665,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.33401283369533513,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.4466,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.37751777334940384,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.4275,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.3220782745130144,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.4443,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.3312790770736199,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.4477,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.3412142240044395,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.4547,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.36324687174387044,
"learning_rate": 2.17791064266051e-05,
"loss": 0.4414,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3315236786143702,
"learning_rate": 2.172276926861741e-05,
"loss": 0.4498,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.371144665634979,
"learning_rate": 2.166641833541255e-05,
"loss": 0.4443,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.3747809566586289,
"learning_rate": 2.161005407757109e-05,
"loss": 0.4133,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.3442494593985286,
"learning_rate": 2.155367694578013e-05,
"loss": 0.4594,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.3811097863331922,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.4382,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.3302155032626096,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.4544,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.35305208850255093,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.4351,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.34781620381859296,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.4474,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.33684452179746216,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.4408,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.38386121280540547,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.4349,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.3600358238205969,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.431,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.34744398923685965,
"learning_rate": 2.110225054402733e-05,
"loss": 0.4549,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.36952093383024864,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.4405,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.3634430215704652,
"learning_rate": 2.09892967807724e-05,
"loss": 0.4616,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3996353416209894,
"learning_rate": 2.093280780774835e-05,
"loss": 0.4155,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.32865607600197616,
"learning_rate": 2.087631137601793e-05,
"loss": 0.4118,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.3788102127685201,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.431,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.345109935908219,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.4497,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.33912933150645025,
"learning_rate": 2.070678184630458e-05,
"loss": 0.4507,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.34981803602142625,
"learning_rate": 2.06502600977305e-05,
"loss": 0.4563,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.38342576644188436,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.4543,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.3912509289941098,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.4408,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.35360331312366355,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.4298,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.37232011470190535,
"learning_rate": 2.042412562887594e-05,
"loss": 0.4317,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.37197267035708664,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.4347,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.322774599833,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.4267,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.38654108861917225,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.4201,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.3326684740577316,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.4287,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.3656895878856609,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.4352,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.34899501051886295,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.4459,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.3726427844252108,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.4327,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.36163722135059023,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.419,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.3683851725431275,
"learning_rate": 1.99151687695709e-05,
"loss": 0.4476,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.3522229659437596,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.4323,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.34834322564929887,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.4557,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.4058350430652889,
"learning_rate": 1.974551241345447e-05,
"loss": 0.4388,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.3316160278411339,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.4411,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.5049791640533998,
"learning_rate": 1.963241759676897e-05,
"loss": 0.4431,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.358078391655462,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.4357,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.3571535163639916,
"learning_rate": 1.951933453677638e-05,
"loss": 0.4175,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.34158188974795944,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.4554,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.3347609676553006,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.4299,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.36623618292713006,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.4368,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.36036611520331274,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.4573,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.38814918575649104,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.4301,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.37736861618309125,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.433,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.42112138695237894,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.4324,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.37849928909838815,
"learning_rate": 1.906719219225165e-05,
"loss": 0.4412,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.39498826196889086,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.4483,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.3742559878981588,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.4288,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.34935944032817323,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.4365,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.3162845791245236,
"learning_rate": 1.884128556891707e-05,
"loss": 0.4455,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.3552709279432287,
"learning_rate": 1.878483094691119e-05,
"loss": 0.4428,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3412753622327309,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.4578,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.3409707555533485,
"learning_rate": 1.867195130360947e-05,
"loss": 0.4225,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.366061144673441,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.4279,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.3335887197962578,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.4576,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.38223458479031697,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.4357,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.3433377112481041,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.4552,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.3542774445435113,
"learning_rate": 1.838994592242891e-05,
"loss": 0.4532,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.34237202365336344,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.442,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.36074536957169123,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.4453,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.31662421845964905,
"learning_rate": 1.82208935733949e-05,
"loss": 0.4408,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.3449507539578164,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.4346,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.355157743850847,
"learning_rate": 1.810826238483892e-05,
"loss": 0.4361,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.32961213410539947,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.4281,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.357129702505585,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.4282,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.32410502087784876,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.443,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.36911260811612706,
"learning_rate": 1.788318512320838e-05,
"loss": 0.4029,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.31095065043294934,
"learning_rate": 1.782695699829051e-05,
"loss": 0.4359,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3528433695821032,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.442,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.3304034188543602,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.4402,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.34067274928895364,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.4523,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.3723257645732867,
"learning_rate": 1.76022227484387e-05,
"loss": 0.4568,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.3472084689416254,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.4367,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.3753486811393218,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.4355,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.34093650344702375,
"learning_rate": 1.743387180005809e-05,
"loss": 0.4428,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.35299756152928485,
"learning_rate": 1.737779525652899e-05,
"loss": 0.4621,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.37426334567574304,
"learning_rate": 1.732173968007914e-05,
"loss": 0.4278,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.3738996964242624,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.4473,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.3562389876662578,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.437,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.33745991628939376,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.4415,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.3620302290891436,
"learning_rate": 1.709773600686015e-05,
"loss": 0.4438,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.34517074355824473,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.4294,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.382605167745748,
"learning_rate": 1.69858716181154e-05,
"loss": 0.426,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.32966817615721783,
"learning_rate": 1.692997535150948e-05,
"loss": 0.4237,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.39743262281020464,
"learning_rate": 1.687410363273884e-05,
"loss": 0.4304,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.35126033492691483,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.4302,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.35458355793156,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.4406,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.400579854047203,
"learning_rate": 1.67066402299235e-05,
"loss": 0.4469,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.3224847550169889,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.4398,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.40288161713270615,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.4307,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.347498788017586,
"learning_rate": 1.653941382839745e-05,
"loss": 0.4219,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.5407597109422304,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.3636,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.3831528287508976,
"learning_rate": 1.642806717162757e-05,
"loss": 0.3666,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.6879965625591378,
"learning_rate": 1.637243646233718e-05,
"loss": 0.3784,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.3885886386210758,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.3234,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.596721665460169,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.3503,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.44880367913411723,
"learning_rate": 1.620572014799777e-05,
"loss": 0.354,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.3971972721622893,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.3453,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.5523594367943542,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.3732,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.42314897922393074,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.3456,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.38681425221643095,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.3372,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.4298015671990596,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.3386,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.42196612597895566,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.3474,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.368089648872561,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.3524,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.4180512183716573,
"learning_rate": 1.576251073999917e-05,
"loss": 0.3402,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.3845286588562857,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.345,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.3766635464598193,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.3531,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.3563091315790123,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.3479,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.35911462817356743,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.3773,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.3608520822145827,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.363,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3245900401407776,
"learning_rate": 1.543151434154755e-05,
"loss": 0.355,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.35392642566867344,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.3607,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.33017431921362367,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.3673,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.3457584142424063,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.3458,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.32530734011326423,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.368,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.34767949003052007,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.3354,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.34714142193768655,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.3629,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.34565581988631583,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.3453,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.36201335982333727,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.3433,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.31998069332391765,
"learning_rate": 1.493751509814374e-05,
"loss": 0.3308,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.3209923666477641,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.3726,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.32654868812346566,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.3717,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.3044880862982792,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.3703,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.31136353377461057,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.389,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.3275248059310185,
"learning_rate": 1.466446738480111e-05,
"loss": 0.3501,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.30069278392713317,
"learning_rate": 1.460998408550112e-05,
"loss": 0.3658,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.32130987844891373,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.3389,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.2995757531371438,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.3625,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.29654957349115624,
"learning_rate": 1.444679451900437e-05,
"loss": 0.3424,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.3272903146209085,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.3546,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.294519442041211,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.3783,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.33507158977650575,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.3644,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.31559653802502374,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.3777,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.3061118087622946,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.3427,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.3189220259107353,
"learning_rate": 1.412162598317602e-05,
"loss": 0.3383,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.2972954355938839,
"learning_rate": 1.406759321421907e-05,
"loss": 0.3529,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.32680925174101166,
"learning_rate": 1.401360788062871e-05,
"loss": 0.3622,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3176941837776654,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.3489,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.2949579138637086,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.3621,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.3370598415692727,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.354,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.3142446487529919,
"learning_rate": 1.379814952747693e-05,
"loss": 0.3385,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.3158293191280785,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.3264,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3104259301113294,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.384,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.3389308191601795,
"learning_rate": 1.363707494698459e-05,
"loss": 0.3331,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.312618377281516,
"learning_rate": 1.358348460369265e-05,
"loss": 0.3686,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.3093720062377724,
"learning_rate": 1.352994556668684e-05,
"loss": 0.3339,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.3262473208459869,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.3695,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.3088344545580555,
"learning_rate": 1.342302312350678e-05,
"loss": 0.3533,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.32019271853417985,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.3625,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.2987412035837677,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.3514,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.31913786400062744,
"learning_rate": 1.326303494478451e-05,
"loss": 0.3322,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.3055864431266222,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.3701,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.30643544234569464,
"learning_rate": 1.315664479122898e-05,
"loss": 0.3609,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.29514619607141335,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.3627,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.311902558517713,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.3279,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.29121678283500846,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.3633,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.28779878837014516,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.3644,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.3261682973980225,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.3431,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.29868853110684934,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.3509,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.31577811166993286,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.3476,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.2906616610742381,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.3753,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.2847926581346661,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.3764,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.3142875216875645,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.356,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.29746653348111984,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.3451,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.3038791235248449,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.3595,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.2988720520010286,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.3602,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.30179084296496084,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.3413,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.28556414431994775,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.351,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.2992643195305742,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.3579,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.29069613625585206,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.3545,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.30460068829754366,
"learning_rate": 1.220938883250288e-05,
"loss": 0.3541,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.3082437011353175,
"learning_rate": 1.215733275116827e-05,
"loss": 0.3541,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.2898395268118843,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.3364,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.303215667233834,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.352,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.2964170277104741,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.3515,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.31063370536973955,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.329,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.2969312987922691,
"learning_rate": 1.189800129803139e-05,
"loss": 0.368,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.28869513442574324,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.3585,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.31176490538241264,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.3423,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.29640418436769744,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.3788,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.2978331701193592,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.3446,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.2913757132476176,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.3532,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.2927151727586002,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.3426,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.3039196261536556,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.3648,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.29373142180264195,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.3525,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.303454838592817,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.3336,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.3110757930940587,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.342,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.29361058933000167,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.3508,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.30852641969856515,
"learning_rate": 1.128231163049258e-05,
"loss": 0.3402,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.29384004711251815,
"learning_rate": 1.123144751066259e-05,
"loss": 0.3608,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.3432049831763918,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.3729,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.31476994725484453,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.3593,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.27513242879156774,
"learning_rate": 1.107927745387114e-05,
"loss": 0.3679,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.2878176238622425,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.3398,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.328133723978726,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.3493,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.30329000787783156,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.368,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.29649664456425734,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.3549,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.29177300493421426,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.3411,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.3224810803476119,
"learning_rate": 1.077687414070869e-05,
"loss": 0.3457,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.2896386589370974,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.3527,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.3726831200559473,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.3833,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.2881050320310152,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.3671,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.28925399781240435,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.354,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.29033426538174195,
"learning_rate": 1.052689572753247e-05,
"loss": 0.3618,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.3010050697975516,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.3513,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.2856259170370127,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.3415,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.30319368519688134,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.3376,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.3149174485439283,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.3631,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.3066390457414162,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.3392,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.31922351989361775,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.3463,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.29932715664448245,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.3451,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.3117407984092045,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.3572,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.28877558797943254,
"learning_rate": 1.008173988718005e-05,
"loss": 0.3597,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.30789366442410687,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.348,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.31544232067158134,
"learning_rate": 9.983678617766677e-06,
"loss": 0.3561,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.2944288929086434,
"learning_rate": 9.934767922553788e-06,
"loss": 0.3441,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.3003814830129868,
"learning_rate": 9.885937708668371e-06,
"loss": 0.3712,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.3060711763732664,
"learning_rate": 9.837188366555524e-06,
"loss": 0.3481,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.29182355433721086,
"learning_rate": 9.788520286013691e-06,
"loss": 0.3587,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.2797735459231335,
"learning_rate": 9.739933856191537e-06,
"loss": 0.3488,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.3027736816340779,
"learning_rate": 9.691429465584866e-06,
"loss": 0.3505,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.32248921055044283,
"learning_rate": 9.643007502033492e-06,
"loss": 0.3584,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.2947915571349239,
"learning_rate": 9.594668352718152e-06,
"loss": 0.3501,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.3219807669538293,
"learning_rate": 9.546412404157383e-06,
"loss": 0.368,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.28842736398112645,
"learning_rate": 9.498240042204471e-06,
"loss": 0.3695,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.29625645472307605,
"learning_rate": 9.450151652044339e-06,
"loss": 0.354,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.3017563158701754,
"learning_rate": 9.402147618190472e-06,
"loss": 0.3761,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.29082524491820105,
"learning_rate": 9.354228324481841e-06,
"loss": 0.3509,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.3054255334092239,
"learning_rate": 9.306394154079852e-06,
"loss": 0.3525,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.37984368463642154,
"learning_rate": 9.258645489465254e-06,
"loss": 0.3518,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.2890949449245077,
"learning_rate": 9.21098271243509e-06,
"loss": 0.3545,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.29595385097074445,
"learning_rate": 9.163406204099659e-06,
"loss": 0.3594,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.3019319322332989,
"learning_rate": 9.115916344879452e-06,
"loss": 0.3653,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.2952317212839934,
"learning_rate": 9.068513514502126e-06,
"loss": 0.3498,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.2906876212695074,
"learning_rate": 9.021198091999439e-06,
"loss": 0.3536,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.29313157669754664,
"learning_rate": 8.973970455704251e-06,
"loss": 0.3501,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.29129413473527244,
"learning_rate": 8.926830983247495e-06,
"loss": 0.3444,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.2877923106148405,
"learning_rate": 8.879780051555116e-06,
"loss": 0.3779,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.30275736892546623,
"learning_rate": 8.83281803684512e-06,
"loss": 0.3371,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.29846600239538545,
"learning_rate": 8.785945314624532e-06,
"loss": 0.3605,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.325584583565924,
"learning_rate": 8.739162259686385e-06,
"loss": 0.3438,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.30874931436962627,
"learning_rate": 8.692469246106734e-06,
"loss": 0.333,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.2855133527879841,
"learning_rate": 8.645866647241677e-06,
"loss": 0.3582,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.3056536510948254,
"learning_rate": 8.599354835724348e-06,
"loss": 0.3341,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.3100776527595224,
"learning_rate": 8.552934183461958e-06,
"loss": 0.3378,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.28090574785516537,
"learning_rate": 8.506605061632805e-06,
"loss": 0.3483,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.30690822333594175,
"learning_rate": 8.460367840683308e-06,
"loss": 0.3328,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.3070425230837834,
"learning_rate": 8.414222890325065e-06,
"loss": 0.3116,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3029867132478452,
"learning_rate": 8.368170579531855e-06,
"loss": 0.3512,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.30435370425525426,
"learning_rate": 8.322211276536733e-06,
"loss": 0.3474,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.30053839602240656,
"learning_rate": 8.27634534882907e-06,
"loss": 0.3434,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.29356415340007475,
"learning_rate": 8.230573163151585e-06,
"loss": 0.3692,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.2786391241230992,
"learning_rate": 8.184895085497466e-06,
"loss": 0.3462,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.29253287901470204,
"learning_rate": 8.139311481107401e-06,
"loss": 0.3592,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.277866520775664,
"learning_rate": 8.093822714466686e-06,
"loss": 0.3424,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.3028304079394389,
"learning_rate": 8.048429149302271e-06,
"loss": 0.3646,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.2847530615696395,
"learning_rate": 8.003131148579902e-06,
"loss": 0.3592,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.2775953557027004,
"learning_rate": 7.957929074501182e-06,
"loss": 0.3514,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.30021767021782964,
"learning_rate": 7.912823288500695e-06,
"loss": 0.341,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.3023581270837931,
"learning_rate": 7.867814151243102e-06,
"loss": 0.3613,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.2702703458106593,
"learning_rate": 7.822902022620267e-06,
"loss": 0.3517,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.29824311590714886,
"learning_rate": 7.778087261748377e-06,
"loss": 0.3449,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.2889624304543533,
"learning_rate": 7.733370226965056e-06,
"loss": 0.34,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.2685899064561315,
"learning_rate": 7.688751275826527e-06,
"loss": 0.3586,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.2836657176610411,
"learning_rate": 7.644230765104735e-06,
"loss": 0.3781,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.29591414598939175,
"learning_rate": 7.599809050784502e-06,
"loss": 0.3558,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.2719275770293297,
"learning_rate": 7.555486488060661e-06,
"loss": 0.3588,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.28253978795551593,
"learning_rate": 7.511263431335249e-06,
"loss": 0.3586,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.29467734119318534,
"learning_rate": 7.467140234214656e-06,
"loss": 0.3606,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.2915565081750334,
"learning_rate": 7.423117249506777e-06,
"loss": 0.3843,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.27979078705365523,
"learning_rate": 7.379194829218231e-06,
"loss": 0.3514,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.2798290993260753,
"learning_rate": 7.335373324551518e-06,
"loss": 0.3676,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.2986744552335149,
"learning_rate": 7.291653085902224e-06,
"loss": 0.3503,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.2798961145408598,
"learning_rate": 7.248034462856213e-06,
"loss": 0.3749,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.2892808193332738,
"learning_rate": 7.20451780418683e-06,
"loss": 0.3424,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.3020778447885371,
"learning_rate": 7.161103457852128e-06,
"loss": 0.3522,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.2896552942230868,
"learning_rate": 7.117791770992046e-06,
"loss": 0.366,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.27730269589860407,
"learning_rate": 7.074583089925688e-06,
"loss": 0.3682,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.27221163322705644,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.3604,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.28447339027371094,
"learning_rate": 6.988476126329604e-06,
"loss": 0.36,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.29473622155965246,
"learning_rate": 6.945578532308857e-06,
"loss": 0.3222,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.2800701820611529,
"learning_rate": 6.902785321094301e-06,
"loss": 0.3407,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.2734736819877084,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.356,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.2965427082920547,
"learning_rate": 6.817513414939892e-06,
"loss": 0.3505,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.2978809486458644,
"learning_rate": 6.775035401831924e-06,
"loss": 0.346,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.29095076565553746,
"learning_rate": 6.732663135188473e-06,
"loss": 0.3341,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.30103298627161645,
"learning_rate": 6.690396953817055e-06,
"loss": 0.365,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.27818940647446705,
"learning_rate": 6.648237195676927e-06,
"loss": 0.3827,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.28516210717929696,
"learning_rate": 6.606184197876395e-06,
"loss": 0.3409,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.31184855218655055,
"learning_rate": 6.564238296670109e-06,
"loss": 0.345,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.28232828120392167,
"learning_rate": 6.52239982745638e-06,
"loss": 0.33,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.2918154648664019,
"learning_rate": 6.480669124774484e-06,
"loss": 0.3735,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.28648832963293874,
"learning_rate": 6.439046522302013e-06,
"loss": 0.3522,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.28520460481198967,
"learning_rate": 6.397532352852191e-06,
"loss": 0.3446,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.30344544231267545,
"learning_rate": 6.356126948371202e-06,
"loss": 0.3343,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.2902668291933707,
"learning_rate": 6.314830639935559e-06,
"loss": 0.3567,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.28547903926342477,
"learning_rate": 6.273643757749448e-06,
"loss": 0.3523,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.3368561209534433,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.3488,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.278223521347598,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.3238,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.2980892569656734,
"learning_rate": 6.150742957589784e-06,
"loss": 0.3556,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.2822058897242162,
"learning_rate": 6.109997064904774e-06,
"loss": 0.352,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.29869861503699446,
"learning_rate": 6.069362236313092e-06,
"loss": 0.3436,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.2748209292394878,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.3671,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.2767705341016716,
"learning_rate": 5.988427070179108e-06,
"loss": 0.3721,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.3046044716158366,
"learning_rate": 5.948127379792234e-06,
"loss": 0.3543,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.29626930198455753,
"learning_rate": 5.90794004780439e-06,
"loss": 0.3522,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.2819924341887773,
"learning_rate": 5.867865395552405e-06,
"loss": 0.3639,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.2979525250916473,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.3607,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.290530148800947,
"learning_rate": 5.788055411095892e-06,
"loss": 0.3331,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.27212710634930076,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.3799,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.2748858403514994,
"learning_rate": 5.708699979051613e-06,
"loss": 0.3621,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.26943876660329535,
"learning_rate": 5.66919351390751e-06,
"loss": 0.3791,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.27688699215792173,
"learning_rate": 5.629801637510188e-06,
"loss": 0.3599,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.2967964313366418,
"learning_rate": 5.590524664836043e-06,
"loss": 0.354,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.275583939693558,
"learning_rate": 5.551362909942708e-06,
"loss": 0.3639,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.2725294260505944,
"learning_rate": 5.51231668596653e-06,
"loss": 0.3568,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.2955650402988491,
"learning_rate": 5.47338630512007e-06,
"loss": 0.3559,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.2793841459407149,
"learning_rate": 5.434572078689617e-06,
"loss": 0.3493,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.2897838815698071,
"learning_rate": 5.395874317032694e-06,
"loss": 0.3653,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.29250686141942117,
"learning_rate": 5.357293329575573e-06,
"loss": 0.3708,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.28494479533295924,
"learning_rate": 5.318829424810797e-06,
"loss": 0.3387,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.27900794469247964,
"learning_rate": 5.280482910294728e-06,
"loss": 0.3469,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.28752279144656534,
"learning_rate": 5.242254092645078e-06,
"loss": 0.327,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.2729039963063428,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.3795,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.26951086733808904,
"learning_rate": 5.166150769707892e-06,
"loss": 0.3746,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.28941035626881406,
"learning_rate": 5.128276872940503e-06,
"loss": 0.3346,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.28207556362433406,
"learning_rate": 5.090521890074953e-06,
"loss": 0.3223,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.2837258509286893,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.3345,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.29290037846973266,
"learning_rate": 5.015369872647442e-06,
"loss": 0.3644,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.2754443981382317,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.3595,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.2739921270170233,
"learning_rate": 4.940697121074675e-06,
"loss": 0.36,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.28990900072156284,
"learning_rate": 4.903541216934864e-06,
"loss": 0.334,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.25700315741255114,
"learning_rate": 4.8665060236772e-06,
"loss": 0.359,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.2758299804444433,
"learning_rate": 4.829591837434102e-06,
"loss": 0.3327,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.28025902836512656,
"learning_rate": 4.792798953370424e-06,
"loss": 0.3308,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.28365442809122693,
"learning_rate": 4.756127665681083e-06,
"loss": 0.3291,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.2766938665962735,
"learning_rate": 4.719578267588725e-06,
"loss": 0.3613,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.2537963246643027,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.3509,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.4184575050925184,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.3121,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.35627457960336956,
"learning_rate": 4.610664328486485e-06,
"loss": 0.2988,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.2700111627500637,
"learning_rate": 4.5746054014809e-06,
"loss": 0.2967,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.36115457496575815,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.2827,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.4867793114707301,
"learning_rate": 4.502857857942204e-06,
"loss": 0.2945,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.3629455664594879,
"learning_rate": 4.467169815100549e-06,
"loss": 0.3028,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.3064647424785411,
"learning_rate": 4.431605972355137e-06,
"loss": 0.3109,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.33607875918969043,
"learning_rate": 4.396166614073516e-06,
"loss": 0.2862,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.3213971062299727,
"learning_rate": 4.360852023627846e-06,
"loss": 0.3177,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.3593199674694445,
"learning_rate": 4.325662483392656e-06,
"loss": 0.3029,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.29323767571647497,
"learning_rate": 4.29059827474257e-06,
"loss": 0.3196,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.26589547661934937,
"learning_rate": 4.255659678050063e-06,
"loss": 0.2864,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.29924678174660113,
"learning_rate": 4.22084697268323e-06,
"loss": 0.2821,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.32812579823109933,
"learning_rate": 4.186160437003535e-06,
"loss": 0.3203,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.32141472957508604,
"learning_rate": 4.151600348363582e-06,
"loss": 0.2983,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.30960561979822077,
"learning_rate": 4.11716698310493e-06,
"loss": 0.2976,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.27399059728016284,
"learning_rate": 4.082860616555848e-06,
"loss": 0.299,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.29824571777441505,
"learning_rate": 4.048681523029125e-06,
"loss": 0.2914,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.3222646718832181,
"learning_rate": 4.014629975819882e-06,
"loss": 0.2927,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.28553145024937304,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.3053,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.2660486762831778,
"learning_rate": 3.94691060843287e-06,
"loss": 0.3028,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.2662352122411146,
"learning_rate": 3.913243329737353e-06,
"loss": 0.2856,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.28538455524816175,
"learning_rate": 3.879704680319503e-06,
"loss": 0.3214,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.26975644636935864,
"learning_rate": 3.846294928353471e-06,
"loss": 0.2954,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.27200212587457295,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.3033,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.2589831184760413,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.2789,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.25855501891074933,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.2888,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.2538923493839949,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.3165,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.24321012644682805,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.3171,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.26491997523571426,
"learning_rate": 3.648558150665886e-06,
"loss": 0.308,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.25313309968660586,
"learning_rate": 3.616058102895075e-06,
"loss": 0.3022,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.2580087235330119,
"learning_rate": 3.583689060687079e-06,
"loss": 0.2949,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.2608350827738475,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.3275,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.25170439397007416,
"learning_rate": 3.51934502719796e-06,
"loss": 0.3151,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.261147183719147,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.2645,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.25862564187004466,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.2774,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.23487887199064736,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.3204,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.25899174739037917,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.2981,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.26569409659567406,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.275,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.24680200834268126,
"learning_rate": 3.32948376099375e-06,
"loss": 0.313,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.24718730486592053,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.2998,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.26352313062463956,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.2863,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.26041978594374365,
"learning_rate": 3.236349233544438e-06,
"loss": 0.2955,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.2502518994712264,
"learning_rate": 3.205572144561604e-06,
"loss": 0.2961,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.25170570999369,
"learning_rate": 3.174929343376374e-06,
"loss": 0.2926,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.25892051069574856,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.3036,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.25417325967151755,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.2819,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.2638129716271809,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.2744,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.25320972774104833,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.291,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.24930614652420457,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.2976,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.26221392788634607,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.2962,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.24591291837446225,
"learning_rate": 2.964210239224696e-06,
"loss": 0.2962,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.2398061848455756,
"learning_rate": 2.934650470499869e-06,
"loss": 0.2938,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.2555839707454249,
"learning_rate": 2.905227155855128e-06,
"loss": 0.2931,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.25026174129964185,
"learning_rate": 2.875940530558519e-06,
"loss": 0.2952,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.240171623818783,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.2758,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.24992413347865844,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.2882,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.24811902997076307,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.2563,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.24642765418379403,
"learning_rate": 2.760165589921184e-06,
"loss": 0.2959,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.26992991450750214,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.2815,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.25956709303477293,
"learning_rate": 2.703104292151164e-06,
"loss": 0.2953,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.24635743573172064,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.3024,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.23734909693237383,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.3114,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.24655558135332595,
"learning_rate": 2.618550200289303e-06,
"loss": 0.2934,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.26181620043895343,
"learning_rate": 2.59064316684412e-06,
"loss": 0.2862,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.2445104429884527,
"learning_rate": 2.562875338152264e-06,
"loss": 0.3116,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.25266410650660565,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.2907,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.25722937332505275,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.291,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.2755383481738704,
"learning_rate": 2.480409295328803e-06,
"loss": 0.315,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.22814028231029848,
"learning_rate": 2.45320049480108e-06,
"loss": 0.3256,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.2468812256707451,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.3174,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.24699626026252963,
"learning_rate": 2.399204021408421e-06,
"loss": 0.2963,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.24513213487556643,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.2793,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.263883652058844,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.3293,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.23962142367619893,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.3052,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.25872281486463067,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.2823,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.25392405143928515,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.3359,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.2562250579123868,
"learning_rate": 2.24059906434998e-06,
"loss": 0.284,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.25132614943363063,
"learning_rate": 2.214660694699091e-06,
"loss": 0.2779,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.2561552339509087,
"learning_rate": 2.188864536147355e-06,
"loss": 0.3157,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.2496355508126803,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.2947,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.25053444189110097,
"learning_rate": 2.137699676264171e-06,
"loss": 0.3056,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2538608557552334,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.2972,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.2436127368671725,
"learning_rate": 2.08710612114857e-06,
"loss": 0.2808,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.24614814912416816,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.2837,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.24965722037475838,
"learning_rate": 2.0370854889762e-06,
"loss": 0.323,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.2331150783100193,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.3399,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.2401363293101567,
"learning_rate": 1.987639379598427e-06,
"loss": 0.2921,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.24208025727791907,
"learning_rate": 1.963132265887919e-06,
"loss": 0.3024,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.24110135467971194,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.3156,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.23617648278912387,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.3044,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.25093902851989347,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.2778,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.24804783446671888,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.3186,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.24254024844945798,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.3108,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.24074601579573965,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.3103,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.2505455067008593,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.2975,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.2350830952363194,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.306,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.24583474536093075,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.3128,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.248097567197119,
"learning_rate": 1.726025117089718e-06,
"loss": 0.2994,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.23570244093663634,
"learning_rate": 1.70311497206598e-06,
"loss": 0.2942,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.2322452707801914,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.3038,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.23192985275055128,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.311,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.24715658216607012,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.3046,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.24929423931761124,
"learning_rate": 1.612939220355272e-06,
"loss": 0.315,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.250700618345655,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.2911,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.25230564984806736,
"learning_rate": 1.568732761074121e-06,
"loss": 0.2699,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.2597092259393831,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.2694,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.24502060793374433,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.3061,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.2483955620432139,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.3027,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.24034978574866367,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.3105,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.260544433046135,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.2737,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.25833325176647004,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.2847,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.23991146236258074,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.2949,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.2543047230116264,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.3027,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.2388464558290493,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.3079,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.24082684226261483,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.2972,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.2442498539720188,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.2931,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.25001065317783483,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.2796,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.24711763307195006,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.2983,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.24615145832447577,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.3035,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.23683257726718515,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.3017,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.25550353580291,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.2909,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.23313413389104332,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.3119,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.23917445579547267,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.3293,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.23124950431006427,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.319,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.24910186383688832,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.282,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.24788728411566768,
"learning_rate": 1.140415161609505e-06,
"loss": 0.2974,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.2524543077656066,
"learning_rate": 1.121666393356604e-06,
"loss": 0.3042,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.24422768719089258,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.3285,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2340471663200146,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.3071,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.249064282017623,
"learning_rate": 1.066326386135097e-06,
"loss": 0.3125,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.23380060482779208,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.3086,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.24606160344122316,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.2561,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2409819494339619,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.2992,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.24222842475401893,
"learning_rate": 9.9465987684501e-07,
"loss": 0.273,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.2638872869659988,
"learning_rate": 9.771228098856534e-07,
"loss": 0.3388,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.2405780788095614,
"learning_rate": 9.597378493454013e-07,
"loss": 0.2994,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.23518918519639737,
"learning_rate": 9.425051342339353e-07,
"loss": 0.3151,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.24724551838635409,
"learning_rate": 9.254248023435864e-07,
"loss": 0.3014,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.24343857907668573,
"learning_rate": 9.084969902482288e-07,
"loss": 0.3067,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.24384291161522417,
"learning_rate": 8.917218333021993e-07,
"loss": 0.2874,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.23421730019531475,
"learning_rate": 8.750994656392109e-07,
"loss": 0.309,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.24576774863865009,
"learning_rate": 8.58630020171265e-07,
"loss": 0.3053,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.24536895076499407,
"learning_rate": 8.423136285876099e-07,
"loss": 0.285,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.240327241011109,
"learning_rate": 8.26150421353682e-07,
"loss": 0.3052,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.24134653329148217,
"learning_rate": 8.101405277100549e-07,
"loss": 0.2816,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.26184806657872206,
"learning_rate": 7.942840756714077e-07,
"loss": 0.3203,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.23843479788569918,
"learning_rate": 7.785811920255093e-07,
"loss": 0.3079,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.23754732864150652,
"learning_rate": 7.630320023322024e-07,
"loss": 0.3151,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.24827288972799888,
"learning_rate": 7.476366309223903e-07,
"loss": 0.3024,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.25476601920269815,
"learning_rate": 7.323952008970537e-07,
"loss": 0.3075,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.2488010250654986,
"learning_rate": 7.173078341262618e-07,
"loss": 0.2958,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.24639501947544284,
"learning_rate": 7.023746512481989e-07,
"loss": 0.2872,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.23415520709226337,
"learning_rate": 6.875957716681902e-07,
"loss": 0.2897,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.24008928460022158,
"learning_rate": 6.729713135577687e-07,
"loss": 0.2892,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.240082259634412,
"learning_rate": 6.585013938537143e-07,
"loss": 0.3045,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.24752628005583727,
"learning_rate": 6.441861282571115e-07,
"loss": 0.3014,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.24730818862002318,
"learning_rate": 6.300256312324471e-07,
"loss": 0.2719,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.24023783593925427,
"learning_rate": 6.160200160066776e-07,
"loss": 0.3166,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.24408690840733233,
"learning_rate": 6.021693945683327e-07,
"loss": 0.2923,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.24961591057840174,
"learning_rate": 5.884738776666044e-07,
"loss": 0.2864,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.24500688162035444,
"learning_rate": 5.749335748104878e-07,
"loss": 0.2916,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.24621894167164154,
"learning_rate": 5.615485942678778e-07,
"loss": 0.3054,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.24053283898069094,
"learning_rate": 5.483190430647289e-07,
"loss": 0.3277,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.2552209763483929,
"learning_rate": 5.352450269841747e-07,
"loss": 0.2913,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.2571212427235213,
"learning_rate": 5.22326650565701e-07,
"loss": 0.2767,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.2288810419956927,
"learning_rate": 5.095640171043048e-07,
"loss": 0.3072,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.2504525075340901,
"learning_rate": 4.969572286496594e-07,
"loss": 0.295,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.25030164883327694,
"learning_rate": 4.845063860053079e-07,
"loss": 0.2794,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.23161903739498213,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.3058,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.2296533159633694,
"learning_rate": 4.600729351261857e-07,
"loss": 0.291,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.2421570995425036,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.2784,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.23993378839931045,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.291,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.23494225343847852,
"learning_rate": 4.245948007320588e-07,
"loss": 0.2949,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.23990931282801076,
"learning_rate": 4.130816799401971e-07,
"loss": 0.3093,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.24013020545097982,
"learning_rate": 4.017251756252827e-07,
"loss": 0.2779,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2398896277225145,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.2899,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.24816368190538618,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.2978,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.22784819994989539,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.2953,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.23426250728946701,
"learning_rate": 3.578671204615347e-07,
"loss": 0.3172,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.22804366073154866,
"learning_rate": 3.472950355547267e-07,
"loss": 0.3324,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.2531977239764307,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.2747,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.2420530867799235,
"learning_rate": 3.266223765343224e-07,
"loss": 0.2846,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.2477599791173756,
"learning_rate": 3.16521967718757e-07,
"loss": 0.2852,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.2459605678901336,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.2956,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.23394834244742496,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.2958,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.24288373236461047,
"learning_rate": 2.871653894244597e-07,
"loss": 0.2952,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.25184188131081237,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.275,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.24711191980697303,
"learning_rate": 2.683823233205485e-07,
"loss": 0.2953,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.2414463806758098,
"learning_rate": 2.592274132703287e-07,
"loss": 0.2812,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.23490865656418566,
"learning_rate": 2.502303499116754e-07,
"loss": 0.2979,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.2279738316286874,
"learning_rate": 2.413912051848666e-07,
"loss": 0.3148,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.24198098907487353,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.3053,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.2365912964903997,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.2923,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.23424890687852867,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.2925,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.23109699088237634,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.33,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.23500700378533393,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.2837,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.24736915144500754,
"learning_rate": 1.916764975761254e-07,
"loss": 0.3068,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.23789440231011336,
"learning_rate": 1.839446914621923e-07,
"loss": 0.3092,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.23125915463658106,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.3135,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.23711927408665745,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.2987,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2574122088535441,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.2912,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.24136679437483521,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.2902,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.24398500194057998,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.2841,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.24320819017303466,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.3087,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.24053030567179648,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.2887,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.24026849067127926,
"learning_rate": 1.277993209368744e-07,
"loss": 0.2916,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.2416218340513298,
"learning_rate": 1.214956261808542e-07,
"loss": 0.2715,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.24071869693865058,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.2886,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.24632010515688774,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.3069,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.23327556158816046,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.2905,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2501460465547567,
"learning_rate": 9.787081204466253e-08,
"loss": 0.3052,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.2314699196515568,
"learning_rate": 9.236233586200627e-08,
"loss": 0.3179,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.24862446551394435,
"learning_rate": 8.70130406196279e-08,
"loss": 0.3197,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.2462992553612317,
"learning_rate": 8.182296909035047e-08,
"loss": 0.322,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.23845008305793977,
"learning_rate": 7.679216277384439e-08,
"loss": 0.2978,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.23598179858303056,
"learning_rate": 7.19206618963053e-08,
"loss": 0.2858,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.23732663921977465,
"learning_rate": 6.720850541012347e-08,
"loss": 0.3258,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.23926116658103497,
"learning_rate": 6.265573099357714e-08,
"loss": 0.3113,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.24286671533258694,
"learning_rate": 5.826237505053067e-08,
"loss": 0.2806,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.23552550730114322,
"learning_rate": 5.4028472710137e-08,
"loss": 0.3088,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.24874537807139296,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.2787,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.2373941426565261,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.2888,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.2333399025080368,
"learning_rate": 4.228381947004812e-08,
"loss": 0.2968,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.24548368142008578,
"learning_rate": 3.868805732811343e-08,
"loss": 0.2893,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.24122613257083594,
"learning_rate": 3.525190530455636e-08,
"loss": 0.2991,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.24126534392148166,
"learning_rate": 3.197539087475887e-08,
"loss": 0.315,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.23634566813907054,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.2853,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.22965814636321424,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.3015,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.23107884673305867,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.2987,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.2535470021221942,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.294,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.22460969890861898,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.3239,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.2442462091298324,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.2919,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.22885323311836767,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.3173,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.23190290510600686,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.3064,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.2597103058160472,
"learning_rate": 9.67435416023843e-09,
"loss": 0.3086,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.2328815268549642,
"learning_rate": 7.995445897566178e-09,
"loss": 0.3374,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.24256317090842186,
"learning_rate": 6.476393170578021e-09,
"loss": 0.2854,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.2479179140741735,
"learning_rate": 5.11720812557348e-09,
"loss": 0.2827,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.2430224738408543,
"learning_rate": 3.917901630563492e-09,
"loss": 0.2967,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.23531062690448765,
"learning_rate": 2.878483275170485e-09,
"loss": 0.304,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.24194671170565651,
"learning_rate": 1.998961370555108e-09,
"loss": 0.2819,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.2475293033286801,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.3001,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.23829192650513042,
"learning_rate": 7.196337656112384e-10,
"loss": 0.3217,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.23464330127082292,
"learning_rate": 3.198382947511292e-10,
"loss": 0.3085,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.2360860489091689,
"learning_rate": 7.995973352770137e-11,
"loss": 0.3146,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.23356212117981287,
"learning_rate": 0.0,
"loss": 0.3072,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 3.233802660553425e+18,
"train_loss": 0.4578841227992826,
"train_runtime": 28049.3395,
"train_samples_per_second": 5.633,
"train_steps_per_second": 0.044
}
],
"logging_steps": 1,
"max_steps": 1235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.233802660553425e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}