Enigma-70B-Entity-Attribute / trainer_state.json
SeacowX's picture
Upload folder using huggingface_hub
669804f verified
Raw
History Blame Contribute Delete
130 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 747,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004016064257028112,
"grad_norm": 5.430129528045654,
"learning_rate": 6.666666666666667e-07,
"loss": 2.9997,
"step": 1
},
{
"epoch": 0.008032128514056224,
"grad_norm": 4.9348368644714355,
"learning_rate": 1.3333333333333334e-06,
"loss": 3.1432,
"step": 2
},
{
"epoch": 0.012048192771084338,
"grad_norm": 5.114020824432373,
"learning_rate": 2.0000000000000003e-06,
"loss": 3.216,
"step": 3
},
{
"epoch": 0.01606425702811245,
"grad_norm": 4.9010186195373535,
"learning_rate": 2.666666666666667e-06,
"loss": 2.867,
"step": 4
},
{
"epoch": 0.020080321285140562,
"grad_norm": 6.092567443847656,
"learning_rate": 3.3333333333333333e-06,
"loss": 3.2748,
"step": 5
},
{
"epoch": 0.024096385542168676,
"grad_norm": 5.692524433135986,
"learning_rate": 4.000000000000001e-06,
"loss": 3.362,
"step": 6
},
{
"epoch": 0.028112449799196786,
"grad_norm": 5.040319919586182,
"learning_rate": 4.666666666666667e-06,
"loss": 3.0622,
"step": 7
},
{
"epoch": 0.0321285140562249,
"grad_norm": 5.454136371612549,
"learning_rate": 5.333333333333334e-06,
"loss": 2.8978,
"step": 8
},
{
"epoch": 0.03614457831325301,
"grad_norm": 5.210033893585205,
"learning_rate": 6e-06,
"loss": 3.2892,
"step": 9
},
{
"epoch": 0.040160642570281124,
"grad_norm": 5.304287433624268,
"learning_rate": 6.666666666666667e-06,
"loss": 3.2337,
"step": 10
},
{
"epoch": 0.04417670682730924,
"grad_norm": 5.1920928955078125,
"learning_rate": 7.333333333333334e-06,
"loss": 2.6911,
"step": 11
},
{
"epoch": 0.04819277108433735,
"grad_norm": 6.0313944816589355,
"learning_rate": 8.000000000000001e-06,
"loss": 3.2769,
"step": 12
},
{
"epoch": 0.05220883534136546,
"grad_norm": 6.19746732711792,
"learning_rate": 8.666666666666668e-06,
"loss": 3.2052,
"step": 13
},
{
"epoch": 0.05622489959839357,
"grad_norm": 5.583286285400391,
"learning_rate": 9.333333333333334e-06,
"loss": 2.7905,
"step": 14
},
{
"epoch": 0.060240963855421686,
"grad_norm": 6.385959148406982,
"learning_rate": 1e-05,
"loss": 2.9643,
"step": 15
},
{
"epoch": 0.0642570281124498,
"grad_norm": 5.707152843475342,
"learning_rate": 1.0666666666666667e-05,
"loss": 2.7474,
"step": 16
},
{
"epoch": 0.06827309236947791,
"grad_norm": 6.302776336669922,
"learning_rate": 1.1333333333333334e-05,
"loss": 2.7656,
"step": 17
},
{
"epoch": 0.07228915662650602,
"grad_norm": 5.756039142608643,
"learning_rate": 1.2e-05,
"loss": 2.4383,
"step": 18
},
{
"epoch": 0.07630522088353414,
"grad_norm": 5.499080181121826,
"learning_rate": 1.2666666666666668e-05,
"loss": 2.422,
"step": 19
},
{
"epoch": 0.08032128514056225,
"grad_norm": 5.025470733642578,
"learning_rate": 1.3333333333333333e-05,
"loss": 2.2713,
"step": 20
},
{
"epoch": 0.08433734939759036,
"grad_norm": 4.855751037597656,
"learning_rate": 1.4000000000000001e-05,
"loss": 2.0279,
"step": 21
},
{
"epoch": 0.08835341365461848,
"grad_norm": 4.215469837188721,
"learning_rate": 1.4666666666666668e-05,
"loss": 1.8403,
"step": 22
},
{
"epoch": 0.09236947791164658,
"grad_norm": 3.61306095123291,
"learning_rate": 1.5333333333333334e-05,
"loss": 1.6641,
"step": 23
},
{
"epoch": 0.0963855421686747,
"grad_norm": 3.166682481765747,
"learning_rate": 1.6000000000000003e-05,
"loss": 1.5952,
"step": 24
},
{
"epoch": 0.10040160642570281,
"grad_norm": 3.2250802516937256,
"learning_rate": 1.6666666666666667e-05,
"loss": 1.5425,
"step": 25
},
{
"epoch": 0.10441767068273092,
"grad_norm": 2.81725811958313,
"learning_rate": 1.7333333333333336e-05,
"loss": 1.3256,
"step": 26
},
{
"epoch": 0.10843373493975904,
"grad_norm": 2.594248056411743,
"learning_rate": 1.8e-05,
"loss": 1.3129,
"step": 27
},
{
"epoch": 0.11244979919678715,
"grad_norm": 2.2820067405700684,
"learning_rate": 1.866666666666667e-05,
"loss": 1.1439,
"step": 28
},
{
"epoch": 0.11646586345381527,
"grad_norm": 1.8870577812194824,
"learning_rate": 1.9333333333333333e-05,
"loss": 1.0066,
"step": 29
},
{
"epoch": 0.12048192771084337,
"grad_norm": 1.5204654932022095,
"learning_rate": 2e-05,
"loss": 0.8846,
"step": 30
},
{
"epoch": 0.12449799196787148,
"grad_norm": 1.673478126525879,
"learning_rate": 2.0666666666666666e-05,
"loss": 0.8376,
"step": 31
},
{
"epoch": 0.1285140562248996,
"grad_norm": 1.2357970476150513,
"learning_rate": 2.1333333333333335e-05,
"loss": 0.6959,
"step": 32
},
{
"epoch": 0.13253012048192772,
"grad_norm": 1.2078135013580322,
"learning_rate": 2.2000000000000003e-05,
"loss": 0.7046,
"step": 33
},
{
"epoch": 0.13654618473895583,
"grad_norm": 0.9034905433654785,
"learning_rate": 2.2666666666666668e-05,
"loss": 0.6082,
"step": 34
},
{
"epoch": 0.14056224899598393,
"grad_norm": 0.7594810128211975,
"learning_rate": 2.3333333333333336e-05,
"loss": 0.5767,
"step": 35
},
{
"epoch": 0.14457831325301204,
"grad_norm": 0.8364936709403992,
"learning_rate": 2.4e-05,
"loss": 0.6495,
"step": 36
},
{
"epoch": 0.14859437751004015,
"grad_norm": 0.8905705809593201,
"learning_rate": 2.466666666666667e-05,
"loss": 0.6567,
"step": 37
},
{
"epoch": 0.15261044176706828,
"grad_norm": 0.6581031680107117,
"learning_rate": 2.5333333333333337e-05,
"loss": 0.5758,
"step": 38
},
{
"epoch": 0.1566265060240964,
"grad_norm": 0.6650463938713074,
"learning_rate": 2.6000000000000002e-05,
"loss": 0.5181,
"step": 39
},
{
"epoch": 0.1606425702811245,
"grad_norm": 0.5890013575553894,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.5293,
"step": 40
},
{
"epoch": 0.1646586345381526,
"grad_norm": 0.7327901124954224,
"learning_rate": 2.733333333333333e-05,
"loss": 0.5551,
"step": 41
},
{
"epoch": 0.1686746987951807,
"grad_norm": 0.5876369476318359,
"learning_rate": 2.8000000000000003e-05,
"loss": 0.6311,
"step": 42
},
{
"epoch": 0.17269076305220885,
"grad_norm": 0.5172324776649475,
"learning_rate": 2.8666666666666668e-05,
"loss": 0.5199,
"step": 43
},
{
"epoch": 0.17670682730923695,
"grad_norm": 0.5804829597473145,
"learning_rate": 2.9333333333333336e-05,
"loss": 0.5169,
"step": 44
},
{
"epoch": 0.18072289156626506,
"grad_norm": 0.5852580666542053,
"learning_rate": 3e-05,
"loss": 0.4837,
"step": 45
},
{
"epoch": 0.18473895582329317,
"grad_norm": 0.5731318593025208,
"learning_rate": 3.066666666666667e-05,
"loss": 0.4394,
"step": 46
},
{
"epoch": 0.18875502008032127,
"grad_norm": 0.6261307001113892,
"learning_rate": 3.1333333333333334e-05,
"loss": 0.4548,
"step": 47
},
{
"epoch": 0.1927710843373494,
"grad_norm": 0.5107020735740662,
"learning_rate": 3.2000000000000005e-05,
"loss": 0.4889,
"step": 48
},
{
"epoch": 0.19678714859437751,
"grad_norm": 0.43027108907699585,
"learning_rate": 3.266666666666667e-05,
"loss": 0.4412,
"step": 49
},
{
"epoch": 0.20080321285140562,
"grad_norm": 0.42177078127861023,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.474,
"step": 50
},
{
"epoch": 0.20481927710843373,
"grad_norm": 0.5765092372894287,
"learning_rate": 3.4000000000000007e-05,
"loss": 0.4871,
"step": 51
},
{
"epoch": 0.20883534136546184,
"grad_norm": 0.6210499405860901,
"learning_rate": 3.466666666666667e-05,
"loss": 0.4179,
"step": 52
},
{
"epoch": 0.21285140562248997,
"grad_norm": 0.914112389087677,
"learning_rate": 3.5333333333333336e-05,
"loss": 0.4378,
"step": 53
},
{
"epoch": 0.21686746987951808,
"grad_norm": 0.4526534080505371,
"learning_rate": 3.6e-05,
"loss": 0.4425,
"step": 54
},
{
"epoch": 0.22088353413654618,
"grad_norm": 0.3655189573764801,
"learning_rate": 3.6666666666666666e-05,
"loss": 0.4721,
"step": 55
},
{
"epoch": 0.2248995983935743,
"grad_norm": 0.5120930075645447,
"learning_rate": 3.733333333333334e-05,
"loss": 0.5405,
"step": 56
},
{
"epoch": 0.2289156626506024,
"grad_norm": 0.6728758811950684,
"learning_rate": 3.8e-05,
"loss": 0.4425,
"step": 57
},
{
"epoch": 0.23293172690763053,
"grad_norm": 0.5174463987350464,
"learning_rate": 3.866666666666667e-05,
"loss": 0.4568,
"step": 58
},
{
"epoch": 0.23694779116465864,
"grad_norm": 0.4026656746864319,
"learning_rate": 3.933333333333333e-05,
"loss": 0.4558,
"step": 59
},
{
"epoch": 0.24096385542168675,
"grad_norm": 0.35401567816734314,
"learning_rate": 4e-05,
"loss": 0.408,
"step": 60
},
{
"epoch": 0.24497991967871485,
"grad_norm": 0.39593473076820374,
"learning_rate": 4.066666666666667e-05,
"loss": 0.4496,
"step": 61
},
{
"epoch": 0.24899598393574296,
"grad_norm": 0.38727280497550964,
"learning_rate": 4.133333333333333e-05,
"loss": 0.4339,
"step": 62
},
{
"epoch": 0.25301204819277107,
"grad_norm": 0.373130738735199,
"learning_rate": 4.2e-05,
"loss": 0.4503,
"step": 63
},
{
"epoch": 0.2570281124497992,
"grad_norm": 0.4998255670070648,
"learning_rate": 4.266666666666667e-05,
"loss": 0.4721,
"step": 64
},
{
"epoch": 0.26104417670682734,
"grad_norm": 0.3641432821750641,
"learning_rate": 4.3333333333333334e-05,
"loss": 0.4205,
"step": 65
},
{
"epoch": 0.26506024096385544,
"grad_norm": 0.5066208243370056,
"learning_rate": 4.4000000000000006e-05,
"loss": 0.4365,
"step": 66
},
{
"epoch": 0.26907630522088355,
"grad_norm": 0.409373939037323,
"learning_rate": 4.466666666666667e-05,
"loss": 0.4603,
"step": 67
},
{
"epoch": 0.27309236947791166,
"grad_norm": 0.34092167019844055,
"learning_rate": 4.5333333333333335e-05,
"loss": 0.3738,
"step": 68
},
{
"epoch": 0.27710843373493976,
"grad_norm": 0.4341434836387634,
"learning_rate": 4.600000000000001e-05,
"loss": 0.453,
"step": 69
},
{
"epoch": 0.28112449799196787,
"grad_norm": 0.43432170152664185,
"learning_rate": 4.666666666666667e-05,
"loss": 0.3984,
"step": 70
},
{
"epoch": 0.285140562248996,
"grad_norm": 0.4759830832481384,
"learning_rate": 4.7333333333333336e-05,
"loss": 0.4755,
"step": 71
},
{
"epoch": 0.2891566265060241,
"grad_norm": 0.44551223516464233,
"learning_rate": 4.8e-05,
"loss": 0.3825,
"step": 72
},
{
"epoch": 0.2931726907630522,
"grad_norm": 0.41405653953552246,
"learning_rate": 4.866666666666667e-05,
"loss": 0.4223,
"step": 73
},
{
"epoch": 0.2971887550200803,
"grad_norm": 0.3871577978134155,
"learning_rate": 4.933333333333334e-05,
"loss": 0.4336,
"step": 74
},
{
"epoch": 0.30120481927710846,
"grad_norm": 0.3929159343242645,
"learning_rate": 5e-05,
"loss": 0.4728,
"step": 75
},
{
"epoch": 0.30522088353413657,
"grad_norm": 0.41161471605300903,
"learning_rate": 4.9999726806462766e-05,
"loss": 0.3749,
"step": 76
},
{
"epoch": 0.3092369477911647,
"grad_norm": 0.3612353503704071,
"learning_rate": 4.9998907231821824e-05,
"loss": 0.4002,
"step": 77
},
{
"epoch": 0.3132530120481928,
"grad_norm": 0.37977099418640137,
"learning_rate": 4.9997541293989384e-05,
"loss": 0.3887,
"step": 78
},
{
"epoch": 0.3172690763052209,
"grad_norm": 0.42254969477653503,
"learning_rate": 4.999562902281866e-05,
"loss": 0.3706,
"step": 79
},
{
"epoch": 0.321285140562249,
"grad_norm": 0.3744674623012543,
"learning_rate": 4.999317046010329e-05,
"loss": 0.3563,
"step": 80
},
{
"epoch": 0.3253012048192771,
"grad_norm": 0.41670969128608704,
"learning_rate": 4.999016565957633e-05,
"loss": 0.3477,
"step": 81
},
{
"epoch": 0.3293172690763052,
"grad_norm": 0.3389250934123993,
"learning_rate": 4.9986614686909146e-05,
"loss": 0.3954,
"step": 82
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.5018749237060547,
"learning_rate": 4.998251761970997e-05,
"loss": 0.3768,
"step": 83
},
{
"epoch": 0.3373493975903614,
"grad_norm": 0.5131321549415588,
"learning_rate": 4.9977874547522175e-05,
"loss": 0.3834,
"step": 84
},
{
"epoch": 0.3413654618473896,
"grad_norm": 0.4155214726924896,
"learning_rate": 4.997268557182235e-05,
"loss": 0.3863,
"step": 85
},
{
"epoch": 0.3453815261044177,
"grad_norm": 0.42339903116226196,
"learning_rate": 4.9966950806018075e-05,
"loss": 0.3708,
"step": 86
},
{
"epoch": 0.3493975903614458,
"grad_norm": 0.36129459738731384,
"learning_rate": 4.996067037544542e-05,
"loss": 0.3387,
"step": 87
},
{
"epoch": 0.3534136546184739,
"grad_norm": 0.6469271183013916,
"learning_rate": 4.995384441736621e-05,
"loss": 0.4066,
"step": 88
},
{
"epoch": 0.357429718875502,
"grad_norm": 0.39704829454421997,
"learning_rate": 4.994647308096509e-05,
"loss": 0.3869,
"step": 89
},
{
"epoch": 0.3614457831325301,
"grad_norm": 0.39340734481811523,
"learning_rate": 4.9938556527346155e-05,
"loss": 0.3688,
"step": 90
},
{
"epoch": 0.3654618473895582,
"grad_norm": 0.3902013599872589,
"learning_rate": 4.9930094929529506e-05,
"loss": 0.3864,
"step": 91
},
{
"epoch": 0.36947791164658633,
"grad_norm": 0.5779062509536743,
"learning_rate": 4.992108847244746e-05,
"loss": 0.3811,
"step": 92
},
{
"epoch": 0.37349397590361444,
"grad_norm": 0.49305233359336853,
"learning_rate": 4.991153735294049e-05,
"loss": 0.5095,
"step": 93
},
{
"epoch": 0.37751004016064255,
"grad_norm": 0.36875924468040466,
"learning_rate": 4.99014417797529e-05,
"loss": 0.3703,
"step": 94
},
{
"epoch": 0.3815261044176707,
"grad_norm": 0.40627428889274597,
"learning_rate": 4.989080197352834e-05,
"loss": 0.3436,
"step": 95
},
{
"epoch": 0.3855421686746988,
"grad_norm": 0.536807119846344,
"learning_rate": 4.987961816680492e-05,
"loss": 0.3215,
"step": 96
},
{
"epoch": 0.3895582329317269,
"grad_norm": 0.41531243920326233,
"learning_rate": 4.9867890604010125e-05,
"loss": 0.4319,
"step": 97
},
{
"epoch": 0.39357429718875503,
"grad_norm": 0.42658236622810364,
"learning_rate": 4.9855619541455514e-05,
"loss": 0.4457,
"step": 98
},
{
"epoch": 0.39759036144578314,
"grad_norm": 0.3963005840778351,
"learning_rate": 4.984280524733107e-05,
"loss": 0.3767,
"step": 99
},
{
"epoch": 0.40160642570281124,
"grad_norm": 0.4746464490890503,
"learning_rate": 4.9829448001699384e-05,
"loss": 0.4118,
"step": 100
},
{
"epoch": 0.40562248995983935,
"grad_norm": 0.4427131712436676,
"learning_rate": 4.981554809648952e-05,
"loss": 0.4204,
"step": 101
},
{
"epoch": 0.40963855421686746,
"grad_norm": 0.5352081060409546,
"learning_rate": 4.980110583549062e-05,
"loss": 0.3458,
"step": 102
},
{
"epoch": 0.41365461847389556,
"grad_norm": 0.5220073461532593,
"learning_rate": 4.9786121534345265e-05,
"loss": 0.4158,
"step": 103
},
{
"epoch": 0.41767068273092367,
"grad_norm": 0.4506036043167114,
"learning_rate": 4.9770595520542594e-05,
"loss": 0.4267,
"step": 104
},
{
"epoch": 0.42168674698795183,
"grad_norm": 0.4048142433166504,
"learning_rate": 4.975452813341114e-05,
"loss": 0.3909,
"step": 105
},
{
"epoch": 0.42570281124497994,
"grad_norm": 0.40757226943969727,
"learning_rate": 4.973791972411141e-05,
"loss": 0.3598,
"step": 106
},
{
"epoch": 0.42971887550200805,
"grad_norm": 0.5002720355987549,
"learning_rate": 4.972077065562821e-05,
"loss": 0.4114,
"step": 107
},
{
"epoch": 0.43373493975903615,
"grad_norm": 0.339193195104599,
"learning_rate": 4.970308130276272e-05,
"loss": 0.4094,
"step": 108
},
{
"epoch": 0.43775100401606426,
"grad_norm": 0.31917133927345276,
"learning_rate": 4.9684852052124285e-05,
"loss": 0.3217,
"step": 109
},
{
"epoch": 0.44176706827309237,
"grad_norm": 0.35490843653678894,
"learning_rate": 4.966608330212198e-05,
"loss": 0.3443,
"step": 110
},
{
"epoch": 0.4457831325301205,
"grad_norm": 0.3730883002281189,
"learning_rate": 4.96467754629559e-05,
"loss": 0.3474,
"step": 111
},
{
"epoch": 0.4497991967871486,
"grad_norm": 0.3151787221431732,
"learning_rate": 4.96269289566082e-05,
"loss": 0.3804,
"step": 112
},
{
"epoch": 0.4538152610441767,
"grad_norm": 0.37523341178894043,
"learning_rate": 4.960654421683386e-05,
"loss": 0.3837,
"step": 113
},
{
"epoch": 0.4578313253012048,
"grad_norm": 0.4448049068450928,
"learning_rate": 4.9585621689151216e-05,
"loss": 0.3886,
"step": 114
},
{
"epoch": 0.46184738955823296,
"grad_norm": 0.44420376420021057,
"learning_rate": 4.956416183083221e-05,
"loss": 0.3524,
"step": 115
},
{
"epoch": 0.46586345381526106,
"grad_norm": 0.3918246030807495,
"learning_rate": 4.9542165110892414e-05,
"loss": 0.4317,
"step": 116
},
{
"epoch": 0.46987951807228917,
"grad_norm": 0.3922871947288513,
"learning_rate": 4.951963201008076e-05,
"loss": 0.4149,
"step": 117
},
{
"epoch": 0.4738955823293173,
"grad_norm": 0.3816840648651123,
"learning_rate": 4.9496563020869064e-05,
"loss": 0.3741,
"step": 118
},
{
"epoch": 0.4779116465863454,
"grad_norm": 0.37001416087150574,
"learning_rate": 4.947295864744121e-05,
"loss": 0.4215,
"step": 119
},
{
"epoch": 0.4819277108433735,
"grad_norm": 0.39867740869522095,
"learning_rate": 4.9448819405682193e-05,
"loss": 0.344,
"step": 120
},
{
"epoch": 0.4859437751004016,
"grad_norm": 0.3945103585720062,
"learning_rate": 4.9424145823166794e-05,
"loss": 0.3441,
"step": 121
},
{
"epoch": 0.4899598393574297,
"grad_norm": 0.4285331070423126,
"learning_rate": 4.9398938439148077e-05,
"loss": 0.3718,
"step": 122
},
{
"epoch": 0.4939759036144578,
"grad_norm": 0.4073355197906494,
"learning_rate": 4.937319780454559e-05,
"loss": 0.3657,
"step": 123
},
{
"epoch": 0.4979919678714859,
"grad_norm": 0.3194061517715454,
"learning_rate": 4.934692448193334e-05,
"loss": 0.3076,
"step": 124
},
{
"epoch": 0.5020080321285141,
"grad_norm": 0.4059401750564575,
"learning_rate": 4.9320119045527487e-05,
"loss": 0.4063,
"step": 125
},
{
"epoch": 0.5060240963855421,
"grad_norm": 0.461825430393219,
"learning_rate": 4.929278208117378e-05,
"loss": 0.3541,
"step": 126
},
{
"epoch": 0.5100401606425703,
"grad_norm": 0.3667549788951874,
"learning_rate": 4.9264914186334775e-05,
"loss": 0.3651,
"step": 127
},
{
"epoch": 0.5140562248995983,
"grad_norm": 0.37440887093544006,
"learning_rate": 4.9236515970076785e-05,
"loss": 0.3213,
"step": 128
},
{
"epoch": 0.5180722891566265,
"grad_norm": 0.3725244402885437,
"learning_rate": 4.9207588053056545e-05,
"loss": 0.3819,
"step": 129
},
{
"epoch": 0.5220883534136547,
"grad_norm": 0.3577054440975189,
"learning_rate": 4.9178131067507625e-05,
"loss": 0.3955,
"step": 130
},
{
"epoch": 0.5261044176706827,
"grad_norm": 0.35495486855506897,
"learning_rate": 4.914814565722671e-05,
"loss": 0.3697,
"step": 131
},
{
"epoch": 0.5301204819277109,
"grad_norm": 0.421565443277359,
"learning_rate": 4.91176324775594e-05,
"loss": 0.3896,
"step": 132
},
{
"epoch": 0.5341365461847389,
"grad_norm": 0.40980878472328186,
"learning_rate": 4.9086592195385974e-05,
"loss": 0.357,
"step": 133
},
{
"epoch": 0.5381526104417671,
"grad_norm": 0.3367171585559845,
"learning_rate": 4.90550254891068e-05,
"loss": 0.347,
"step": 134
},
{
"epoch": 0.5421686746987951,
"grad_norm": 0.4583292603492737,
"learning_rate": 4.9022933048627496e-05,
"loss": 0.3909,
"step": 135
},
{
"epoch": 0.5461847389558233,
"grad_norm": 0.3414941430091858,
"learning_rate": 4.899031557534383e-05,
"loss": 0.3606,
"step": 136
},
{
"epoch": 0.5502008032128514,
"grad_norm": 0.3588523268699646,
"learning_rate": 4.895717378212644e-05,
"loss": 0.3406,
"step": 137
},
{
"epoch": 0.5542168674698795,
"grad_norm": 0.42519375681877136,
"learning_rate": 4.892350839330522e-05,
"loss": 0.3455,
"step": 138
},
{
"epoch": 0.5582329317269076,
"grad_norm": 0.41246649622917175,
"learning_rate": 4.888932014465352e-05,
"loss": 0.3738,
"step": 139
},
{
"epoch": 0.5622489959839357,
"grad_norm": 0.37443602085113525,
"learning_rate": 4.8854609783372014e-05,
"loss": 0.2889,
"step": 140
},
{
"epoch": 0.5662650602409639,
"grad_norm": 0.4313961863517761,
"learning_rate": 4.881937806807241e-05,
"loss": 0.3055,
"step": 141
},
{
"epoch": 0.570281124497992,
"grad_norm": 0.36356812715530396,
"learning_rate": 4.878362576876087e-05,
"loss": 0.339,
"step": 142
},
{
"epoch": 0.5742971887550201,
"grad_norm": 0.5344200134277344,
"learning_rate": 4.874735366682115e-05,
"loss": 0.3456,
"step": 143
},
{
"epoch": 0.5783132530120482,
"grad_norm": 0.41200292110443115,
"learning_rate": 4.871056255499757e-05,
"loss": 0.323,
"step": 144
},
{
"epoch": 0.5823293172690763,
"grad_norm": 0.4514438509941101,
"learning_rate": 4.867325323737765e-05,
"loss": 0.3925,
"step": 145
},
{
"epoch": 0.5863453815261044,
"grad_norm": 0.4414016008377075,
"learning_rate": 4.863542652937453e-05,
"loss": 0.4399,
"step": 146
},
{
"epoch": 0.5903614457831325,
"grad_norm": 0.401784747838974,
"learning_rate": 4.8597083257709194e-05,
"loss": 0.3259,
"step": 147
},
{
"epoch": 0.5943775100401606,
"grad_norm": 0.3991292119026184,
"learning_rate": 4.855822426039236e-05,
"loss": 0.371,
"step": 148
},
{
"epoch": 0.5983935742971888,
"grad_norm": 0.428716242313385,
"learning_rate": 4.851885038670618e-05,
"loss": 0.3827,
"step": 149
},
{
"epoch": 0.6024096385542169,
"grad_norm": 0.3733920753002167,
"learning_rate": 4.84789624971857e-05,
"loss": 0.3192,
"step": 150
},
{
"epoch": 0.606425702811245,
"grad_norm": 0.36820462346076965,
"learning_rate": 4.8438561463599984e-05,
"loss": 0.3669,
"step": 151
},
{
"epoch": 0.6104417670682731,
"grad_norm": 0.3932380676269531,
"learning_rate": 4.839764816893315e-05,
"loss": 0.3469,
"step": 152
},
{
"epoch": 0.6144578313253012,
"grad_norm": 0.42645588517189026,
"learning_rate": 4.8356223507364996e-05,
"loss": 0.5073,
"step": 153
},
{
"epoch": 0.6184738955823293,
"grad_norm": 0.428401380777359,
"learning_rate": 4.831428838425153e-05,
"loss": 0.3279,
"step": 154
},
{
"epoch": 0.6224899598393574,
"grad_norm": 0.44924989342689514,
"learning_rate": 4.827184371610511e-05,
"loss": 0.3894,
"step": 155
},
{
"epoch": 0.6265060240963856,
"grad_norm": 0.39745035767555237,
"learning_rate": 4.822889043057446e-05,
"loss": 0.3327,
"step": 156
},
{
"epoch": 0.6305220883534136,
"grad_norm": 0.34724128246307373,
"learning_rate": 4.818542946642438e-05,
"loss": 0.3272,
"step": 157
},
{
"epoch": 0.6345381526104418,
"grad_norm": 0.3559289872646332,
"learning_rate": 4.814146177351523e-05,
"loss": 0.2988,
"step": 158
},
{
"epoch": 0.6385542168674698,
"grad_norm": 0.4157481789588928,
"learning_rate": 4.8096988312782174e-05,
"loss": 0.4179,
"step": 159
},
{
"epoch": 0.642570281124498,
"grad_norm": 0.41202324628829956,
"learning_rate": 4.805201005621418e-05,
"loss": 0.4387,
"step": 160
},
{
"epoch": 0.6465863453815262,
"grad_norm": 0.420522540807724,
"learning_rate": 4.800652798683277e-05,
"loss": 0.4313,
"step": 161
},
{
"epoch": 0.6506024096385542,
"grad_norm": 0.3893505036830902,
"learning_rate": 4.796054309867053e-05,
"loss": 0.3366,
"step": 162
},
{
"epoch": 0.6546184738955824,
"grad_norm": 0.4160376489162445,
"learning_rate": 4.791405639674941e-05,
"loss": 0.3784,
"step": 163
},
{
"epoch": 0.6586345381526104,
"grad_norm": 0.47081175446510315,
"learning_rate": 4.786706889705872e-05,
"loss": 0.3517,
"step": 164
},
{
"epoch": 0.6626506024096386,
"grad_norm": 0.37175479531288147,
"learning_rate": 4.781958162653297e-05,
"loss": 0.3345,
"step": 165
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.36895692348480225,
"learning_rate": 4.7771595623029394e-05,
"loss": 0.3274,
"step": 166
},
{
"epoch": 0.6706827309236948,
"grad_norm": 0.3520839214324951,
"learning_rate": 4.7723111935305274e-05,
"loss": 0.2788,
"step": 167
},
{
"epoch": 0.6746987951807228,
"grad_norm": 0.3904466927051544,
"learning_rate": 4.767413162299501e-05,
"loss": 0.3915,
"step": 168
},
{
"epoch": 0.678714859437751,
"grad_norm": 0.39216259121894836,
"learning_rate": 4.7624655756587e-05,
"loss": 0.3354,
"step": 169
},
{
"epoch": 0.6827309236947792,
"grad_norm": 0.6722985506057739,
"learning_rate": 4.757468541740019e-05,
"loss": 0.3945,
"step": 170
},
{
"epoch": 0.6867469879518072,
"grad_norm": 0.4317608177661896,
"learning_rate": 4.752422169756048e-05,
"loss": 0.315,
"step": 171
},
{
"epoch": 0.6907630522088354,
"grad_norm": 0.3866491913795471,
"learning_rate": 4.747326569997684e-05,
"loss": 0.3077,
"step": 172
},
{
"epoch": 0.6947791164658634,
"grad_norm": 0.4831099510192871,
"learning_rate": 4.742181853831721e-05,
"loss": 0.3959,
"step": 173
},
{
"epoch": 0.6987951807228916,
"grad_norm": 0.45947879552841187,
"learning_rate": 4.7369881336984153e-05,
"loss": 0.4242,
"step": 174
},
{
"epoch": 0.7028112449799196,
"grad_norm": 0.4758150577545166,
"learning_rate": 4.731745523109029e-05,
"loss": 0.4127,
"step": 175
},
{
"epoch": 0.7068273092369478,
"grad_norm": 0.4288157522678375,
"learning_rate": 4.7264541366433494e-05,
"loss": 0.3533,
"step": 176
},
{
"epoch": 0.7108433734939759,
"grad_norm": 0.34265485405921936,
"learning_rate": 4.721114089947181e-05,
"loss": 0.3225,
"step": 177
},
{
"epoch": 0.714859437751004,
"grad_norm": 0.4539063572883606,
"learning_rate": 4.715725499729826e-05,
"loss": 0.3832,
"step": 178
},
{
"epoch": 0.7188755020080321,
"grad_norm": 0.40317919850349426,
"learning_rate": 4.7102884837615244e-05,
"loss": 0.3729,
"step": 179
},
{
"epoch": 0.7228915662650602,
"grad_norm": 0.44621387124061584,
"learning_rate": 4.7048031608708876e-05,
"loss": 0.2994,
"step": 180
},
{
"epoch": 0.7269076305220884,
"grad_norm": 0.4565678834915161,
"learning_rate": 4.699269650942296e-05,
"loss": 0.4456,
"step": 181
},
{
"epoch": 0.7309236947791165,
"grad_norm": 0.3770541250705719,
"learning_rate": 4.693688074913282e-05,
"loss": 0.364,
"step": 182
},
{
"epoch": 0.7349397590361446,
"grad_norm": 0.3385080397129059,
"learning_rate": 4.6880585547718845e-05,
"loss": 0.3063,
"step": 183
},
{
"epoch": 0.7389558232931727,
"grad_norm": 0.39325591921806335,
"learning_rate": 4.682381213553986e-05,
"loss": 0.3366,
"step": 184
},
{
"epoch": 0.7429718875502008,
"grad_norm": 0.44467633962631226,
"learning_rate": 4.676656175340621e-05,
"loss": 0.3501,
"step": 185
},
{
"epoch": 0.7469879518072289,
"grad_norm": 0.36168763041496277,
"learning_rate": 4.670883565255264e-05,
"loss": 0.3319,
"step": 186
},
{
"epoch": 0.751004016064257,
"grad_norm": 1.0995522737503052,
"learning_rate": 4.665063509461097e-05,
"loss": 0.3623,
"step": 187
},
{
"epoch": 0.7550200803212851,
"grad_norm": 0.3918020725250244,
"learning_rate": 4.65919613515825e-05,
"loss": 0.3416,
"step": 188
},
{
"epoch": 0.7590361445783133,
"grad_norm": 0.4472026228904724,
"learning_rate": 4.653281570581023e-05,
"loss": 0.3927,
"step": 189
},
{
"epoch": 0.7630522088353414,
"grad_norm": 0.42026591300964355,
"learning_rate": 4.64731994499508e-05,
"loss": 0.3658,
"step": 190
},
{
"epoch": 0.7670682730923695,
"grad_norm": 0.36812499165534973,
"learning_rate": 4.641311388694629e-05,
"loss": 0.3139,
"step": 191
},
{
"epoch": 0.7710843373493976,
"grad_norm": 0.35765761137008667,
"learning_rate": 4.6352560329995686e-05,
"loss": 0.318,
"step": 192
},
{
"epoch": 0.7751004016064257,
"grad_norm": 0.36215129494667053,
"learning_rate": 4.6291540102526235e-05,
"loss": 0.3344,
"step": 193
},
{
"epoch": 0.7791164658634538,
"grad_norm": 0.3706530034542084,
"learning_rate": 4.6230054538164474e-05,
"loss": 0.3077,
"step": 194
},
{
"epoch": 0.7831325301204819,
"grad_norm": 0.3829696774482727,
"learning_rate": 4.6168104980707107e-05,
"loss": 0.4035,
"step": 195
},
{
"epoch": 0.7871485943775101,
"grad_norm": 0.38266220688819885,
"learning_rate": 4.6105692784091636e-05,
"loss": 0.3156,
"step": 196
},
{
"epoch": 0.7911646586345381,
"grad_norm": 0.40088626742362976,
"learning_rate": 4.6042819312366755e-05,
"loss": 0.3606,
"step": 197
},
{
"epoch": 0.7951807228915663,
"grad_norm": 0.38621973991394043,
"learning_rate": 4.597948593966256e-05,
"loss": 0.2774,
"step": 198
},
{
"epoch": 0.7991967871485943,
"grad_norm": 0.4210202395915985,
"learning_rate": 4.591569405016049e-05,
"loss": 0.3532,
"step": 199
},
{
"epoch": 0.8032128514056225,
"grad_norm": 0.3344639539718628,
"learning_rate": 4.585144503806312e-05,
"loss": 0.3641,
"step": 200
},
{
"epoch": 0.8072289156626506,
"grad_norm": 0.41193318367004395,
"learning_rate": 4.5786740307563636e-05,
"loss": 0.3473,
"step": 201
},
{
"epoch": 0.8112449799196787,
"grad_norm": 0.38335129618644714,
"learning_rate": 4.5721581272815164e-05,
"loss": 0.3076,
"step": 202
},
{
"epoch": 0.8152610441767069,
"grad_norm": 0.3907615840435028,
"learning_rate": 4.5655969357899874e-05,
"loss": 0.3529,
"step": 203
},
{
"epoch": 0.8192771084337349,
"grad_norm": 0.3504474461078644,
"learning_rate": 4.558990599679787e-05,
"loss": 0.326,
"step": 204
},
{
"epoch": 0.8232931726907631,
"grad_norm": 0.4119699001312256,
"learning_rate": 4.552339263335581e-05,
"loss": 0.3589,
"step": 205
},
{
"epoch": 0.8273092369477911,
"grad_norm": 0.3600112497806549,
"learning_rate": 4.545643072125538e-05,
"loss": 0.3024,
"step": 206
},
{
"epoch": 0.8313253012048193,
"grad_norm": 0.4069507420063019,
"learning_rate": 4.538902172398151e-05,
"loss": 0.3135,
"step": 207
},
{
"epoch": 0.8353413654618473,
"grad_norm": 0.34810563921928406,
"learning_rate": 4.5321167114790385e-05,
"loss": 0.3451,
"step": 208
},
{
"epoch": 0.8393574297188755,
"grad_norm": 0.3801853656768799,
"learning_rate": 4.525286837667727e-05,
"loss": 0.3253,
"step": 209
},
{
"epoch": 0.8433734939759037,
"grad_norm": 0.4279358386993408,
"learning_rate": 4.518412700234406e-05,
"loss": 0.3147,
"step": 210
},
{
"epoch": 0.8473895582329317,
"grad_norm": 0.3296758234500885,
"learning_rate": 4.511494449416671e-05,
"loss": 0.2706,
"step": 211
},
{
"epoch": 0.8514056224899599,
"grad_norm": 0.3898089826107025,
"learning_rate": 4.504532236416234e-05,
"loss": 0.3197,
"step": 212
},
{
"epoch": 0.8554216867469879,
"grad_norm": 0.43563321232795715,
"learning_rate": 4.497526213395623e-05,
"loss": 0.336,
"step": 213
},
{
"epoch": 0.8594377510040161,
"grad_norm": 0.3690038323402405,
"learning_rate": 4.490476533474855e-05,
"loss": 0.3376,
"step": 214
},
{
"epoch": 0.8634538152610441,
"grad_norm": 0.48120006918907166,
"learning_rate": 4.4833833507280884e-05,
"loss": 0.3789,
"step": 215
},
{
"epoch": 0.8674698795180723,
"grad_norm": 0.33086755871772766,
"learning_rate": 4.476246820180259e-05,
"loss": 0.3188,
"step": 216
},
{
"epoch": 0.8714859437751004,
"grad_norm": 0.40550822019577026,
"learning_rate": 4.4690670978036886e-05,
"loss": 0.3791,
"step": 217
},
{
"epoch": 0.8755020080321285,
"grad_norm": 0.3479694724082947,
"learning_rate": 4.461844340514678e-05,
"loss": 0.344,
"step": 218
},
{
"epoch": 0.8795180722891566,
"grad_norm": 0.33544865250587463,
"learning_rate": 4.454578706170075e-05,
"loss": 0.2766,
"step": 219
},
{
"epoch": 0.8835341365461847,
"grad_norm": 0.7442018985748291,
"learning_rate": 4.447270353563828e-05,
"loss": 0.355,
"step": 220
},
{
"epoch": 0.8875502008032129,
"grad_norm": 0.4419419765472412,
"learning_rate": 4.439919442423513e-05,
"loss": 0.3275,
"step": 221
},
{
"epoch": 0.891566265060241,
"grad_norm": 0.31490767002105713,
"learning_rate": 4.4325261334068426e-05,
"loss": 0.3389,
"step": 222
},
{
"epoch": 0.8955823293172691,
"grad_norm": 0.48634758591651917,
"learning_rate": 4.4250905880981574e-05,
"loss": 0.4238,
"step": 223
},
{
"epoch": 0.8995983935742972,
"grad_norm": 0.4701390564441681,
"learning_rate": 4.417612969004891e-05,
"loss": 0.376,
"step": 224
},
{
"epoch": 0.9036144578313253,
"grad_norm": 0.4378860294818878,
"learning_rate": 4.410093439554019e-05,
"loss": 0.349,
"step": 225
},
{
"epoch": 0.9076305220883534,
"grad_norm": 0.5029141306877136,
"learning_rate": 4.4025321640884905e-05,
"loss": 0.2902,
"step": 226
},
{
"epoch": 0.9116465863453815,
"grad_norm": 0.3482571542263031,
"learning_rate": 4.394929307863633e-05,
"loss": 0.3098,
"step": 227
},
{
"epoch": 0.9156626506024096,
"grad_norm": 0.3532014489173889,
"learning_rate": 4.38728503704354e-05,
"loss": 0.3319,
"step": 228
},
{
"epoch": 0.9196787148594378,
"grad_norm": 0.3720119595527649,
"learning_rate": 4.379599518697444e-05,
"loss": 0.3207,
"step": 229
},
{
"epoch": 0.9236947791164659,
"grad_norm": 0.45317885279655457,
"learning_rate": 4.3718729207960586e-05,
"loss": 0.2996,
"step": 230
},
{
"epoch": 0.927710843373494,
"grad_norm": 0.4399602711200714,
"learning_rate": 4.364105412207914e-05,
"loss": 0.3764,
"step": 231
},
{
"epoch": 0.9317269076305221,
"grad_norm": 0.36520886421203613,
"learning_rate": 4.356297162695662e-05,
"loss": 0.3485,
"step": 232
},
{
"epoch": 0.9357429718875502,
"grad_norm": 0.3675477206707001,
"learning_rate": 4.3484483429123656e-05,
"loss": 0.3043,
"step": 233
},
{
"epoch": 0.9397590361445783,
"grad_norm": 0.43594786524772644,
"learning_rate": 4.3405591243977736e-05,
"loss": 0.3872,
"step": 234
},
{
"epoch": 0.9437751004016064,
"grad_norm": 0.3960169851779938,
"learning_rate": 4.332629679574566e-05,
"loss": 0.3169,
"step": 235
},
{
"epoch": 0.9477911646586346,
"grad_norm": 0.3947587311267853,
"learning_rate": 4.324660181744589e-05,
"loss": 0.3667,
"step": 236
},
{
"epoch": 0.9518072289156626,
"grad_norm": 0.4007823169231415,
"learning_rate": 4.316650805085068e-05,
"loss": 0.3561,
"step": 237
},
{
"epoch": 0.9558232931726908,
"grad_norm": 0.37169793248176575,
"learning_rate": 4.308601724644797e-05,
"loss": 0.3309,
"step": 238
},
{
"epoch": 0.9598393574297188,
"grad_norm": 0.3675639033317566,
"learning_rate": 4.300513116340317e-05,
"loss": 0.3688,
"step": 239
},
{
"epoch": 0.963855421686747,
"grad_norm": 0.39086320996284485,
"learning_rate": 4.2923851569520685e-05,
"loss": 0.2691,
"step": 240
},
{
"epoch": 0.9678714859437751,
"grad_norm": 0.34563392400741577,
"learning_rate": 4.28421802412053e-05,
"loss": 0.3328,
"step": 241
},
{
"epoch": 0.9718875502008032,
"grad_norm": 0.43469658493995667,
"learning_rate": 4.276011896342336e-05,
"loss": 0.3861,
"step": 242
},
{
"epoch": 0.9759036144578314,
"grad_norm": 0.3810824155807495,
"learning_rate": 4.267766952966369e-05,
"loss": 0.293,
"step": 243
},
{
"epoch": 0.9799196787148594,
"grad_norm": 0.45030152797698975,
"learning_rate": 4.259483374189851e-05,
"loss": 0.3045,
"step": 244
},
{
"epoch": 0.9839357429718876,
"grad_norm": 0.39684751629829407,
"learning_rate": 4.251161341054396e-05,
"loss": 0.3644,
"step": 245
},
{
"epoch": 0.9879518072289156,
"grad_norm": 0.4050283432006836,
"learning_rate": 4.242801035442058e-05,
"loss": 0.373,
"step": 246
},
{
"epoch": 0.9919678714859438,
"grad_norm": 0.3969316780567169,
"learning_rate": 4.2344026400713545e-05,
"loss": 0.2957,
"step": 247
},
{
"epoch": 0.9959839357429718,
"grad_norm": 0.412476122379303,
"learning_rate": 4.225966338493271e-05,
"loss": 0.2805,
"step": 248
},
{
"epoch": 1.0,
"grad_norm": 0.4108029305934906,
"learning_rate": 4.2174923150872544e-05,
"loss": 0.3023,
"step": 249
},
{
"epoch": 1.0040160642570282,
"grad_norm": 0.3977179527282715,
"learning_rate": 4.208980755057178e-05,
"loss": 0.4002,
"step": 250
},
{
"epoch": 1.0080321285140563,
"grad_norm": 0.33583542704582214,
"learning_rate": 4.2004318444272985e-05,
"loss": 0.3088,
"step": 251
},
{
"epoch": 1.0120481927710843,
"grad_norm": 0.38322827219963074,
"learning_rate": 4.1918457700381855e-05,
"loss": 0.3724,
"step": 252
},
{
"epoch": 1.0160642570281124,
"grad_norm": 0.3357059955596924,
"learning_rate": 4.183222719542643e-05,
"loss": 0.3046,
"step": 253
},
{
"epoch": 1.0200803212851406,
"grad_norm": 0.3509674370288849,
"learning_rate": 4.174562881401602e-05,
"loss": 0.2623,
"step": 254
},
{
"epoch": 1.0240963855421688,
"grad_norm": 0.3359556496143341,
"learning_rate": 4.16586644488001e-05,
"loss": 0.2781,
"step": 255
},
{
"epoch": 1.0281124497991967,
"grad_norm": 0.38229525089263916,
"learning_rate": 4.157133600042686e-05,
"loss": 0.3679,
"step": 256
},
{
"epoch": 1.0321285140562249,
"grad_norm": 0.40952321887016296,
"learning_rate": 4.148364537750172e-05,
"loss": 0.3323,
"step": 257
},
{
"epoch": 1.036144578313253,
"grad_norm": 0.32103827595710754,
"learning_rate": 4.139559449654561e-05,
"loss": 0.3149,
"step": 258
},
{
"epoch": 1.0401606425702812,
"grad_norm": 0.34906044602394104,
"learning_rate": 4.130718528195303e-05,
"loss": 0.3048,
"step": 259
},
{
"epoch": 1.0441767068273093,
"grad_norm": 0.3745776116847992,
"learning_rate": 4.1218419665950094e-05,
"loss": 0.3144,
"step": 260
},
{
"epoch": 1.0481927710843373,
"grad_norm": 0.40731281042099,
"learning_rate": 4.1129299588552193e-05,
"loss": 0.3221,
"step": 261
},
{
"epoch": 1.0522088353413654,
"grad_norm": 0.34410983324050903,
"learning_rate": 4.103982699752167e-05,
"loss": 0.3113,
"step": 262
},
{
"epoch": 1.0562248995983936,
"grad_norm": 0.3899206817150116,
"learning_rate": 4.095000384832522e-05,
"loss": 0.3555,
"step": 263
},
{
"epoch": 1.0602409638554218,
"grad_norm": 0.4419567883014679,
"learning_rate": 4.085983210409114e-05,
"loss": 0.3634,
"step": 264
},
{
"epoch": 1.0642570281124497,
"grad_norm": 0.4059004485607147,
"learning_rate": 4.076931373556646e-05,
"loss": 0.2772,
"step": 265
},
{
"epoch": 1.0682730923694779,
"grad_norm": 0.3983324468135834,
"learning_rate": 4.067845072107383e-05,
"loss": 0.3465,
"step": 266
},
{
"epoch": 1.072289156626506,
"grad_norm": 0.4240141212940216,
"learning_rate": 4.058724504646834e-05,
"loss": 0.3193,
"step": 267
},
{
"epoch": 1.0763052208835342,
"grad_norm": 0.4722084701061249,
"learning_rate": 4.0495698705094044e-05,
"loss": 0.3743,
"step": 268
},
{
"epoch": 1.0803212851405624,
"grad_norm": 0.4552375078201294,
"learning_rate": 4.0403813697740446e-05,
"loss": 0.3746,
"step": 269
},
{
"epoch": 1.0843373493975903,
"grad_norm": 0.42489174008369446,
"learning_rate": 4.0311592032598754e-05,
"loss": 0.3522,
"step": 270
},
{
"epoch": 1.0883534136546185,
"grad_norm": 0.358661413192749,
"learning_rate": 4.021903572521802e-05,
"loss": 0.3159,
"step": 271
},
{
"epoch": 1.0923694779116466,
"grad_norm": 0.3610284626483917,
"learning_rate": 4.012614679846103e-05,
"loss": 0.337,
"step": 272
},
{
"epoch": 1.0963855421686748,
"grad_norm": 0.395988792181015,
"learning_rate": 4.0032927282460146e-05,
"loss": 0.2843,
"step": 273
},
{
"epoch": 1.1004016064257027,
"grad_norm": 0.36076489090919495,
"learning_rate": 3.9939379214572915e-05,
"loss": 0.3099,
"step": 274
},
{
"epoch": 1.104417670682731,
"grad_norm": 0.3546864092350006,
"learning_rate": 3.9845504639337535e-05,
"loss": 0.2825,
"step": 275
},
{
"epoch": 1.108433734939759,
"grad_norm": 0.35579153895378113,
"learning_rate": 3.9751305608428205e-05,
"loss": 0.3141,
"step": 276
},
{
"epoch": 1.1124497991967872,
"grad_norm": 0.3487059772014618,
"learning_rate": 3.965678418061023e-05,
"loss": 0.2768,
"step": 277
},
{
"epoch": 1.1164658634538154,
"grad_norm": 0.34115543961524963,
"learning_rate": 3.956194242169506e-05,
"loss": 0.2565,
"step": 278
},
{
"epoch": 1.1204819277108433,
"grad_norm": 0.5303230285644531,
"learning_rate": 3.946678240449515e-05,
"loss": 0.3491,
"step": 279
},
{
"epoch": 1.1244979919678715,
"grad_norm": 0.38421887159347534,
"learning_rate": 3.937130620877863e-05,
"loss": 0.339,
"step": 280
},
{
"epoch": 1.1285140562248996,
"grad_norm": 0.4269844591617584,
"learning_rate": 3.927551592122388e-05,
"loss": 0.3895,
"step": 281
},
{
"epoch": 1.1325301204819278,
"grad_norm": 0.36492249369621277,
"learning_rate": 3.9179413635373897e-05,
"loss": 0.3307,
"step": 282
},
{
"epoch": 1.1365461847389557,
"grad_norm": 0.3604801893234253,
"learning_rate": 3.908300145159055e-05,
"loss": 0.2495,
"step": 283
},
{
"epoch": 1.140562248995984,
"grad_norm": 0.4078293740749359,
"learning_rate": 3.898628147700868e-05,
"loss": 0.3299,
"step": 284
},
{
"epoch": 1.144578313253012,
"grad_norm": 0.4020484983921051,
"learning_rate": 3.888925582549006e-05,
"loss": 0.3154,
"step": 285
},
{
"epoch": 1.1485943775100402,
"grad_norm": 0.40106770396232605,
"learning_rate": 3.8791926617577144e-05,
"loss": 0.2955,
"step": 286
},
{
"epoch": 1.1526104417670684,
"grad_norm": 0.43215692043304443,
"learning_rate": 3.869429598044679e-05,
"loss": 0.3371,
"step": 287
},
{
"epoch": 1.1566265060240963,
"grad_norm": 0.4728745222091675,
"learning_rate": 3.859636604786372e-05,
"loss": 0.2671,
"step": 288
},
{
"epoch": 1.1606425702811245,
"grad_norm": 0.3952246308326721,
"learning_rate": 3.8498138960133917e-05,
"loss": 0.2299,
"step": 289
},
{
"epoch": 1.1646586345381527,
"grad_norm": 0.38174968957901,
"learning_rate": 3.8399616864057816e-05,
"loss": 0.3077,
"step": 290
},
{
"epoch": 1.1686746987951806,
"grad_norm": 0.37558120489120483,
"learning_rate": 3.830080191288342e-05,
"loss": 0.3421,
"step": 291
},
{
"epoch": 1.1726907630522088,
"grad_norm": 0.4501299560070038,
"learning_rate": 3.82016962662592e-05,
"loss": 0.3609,
"step": 292
},
{
"epoch": 1.176706827309237,
"grad_norm": 0.35799673199653625,
"learning_rate": 3.810230209018694e-05,
"loss": 0.2487,
"step": 293
},
{
"epoch": 1.180722891566265,
"grad_norm": 0.4033254384994507,
"learning_rate": 3.8002621556974367e-05,
"loss": 0.2843,
"step": 294
},
{
"epoch": 1.1847389558232932,
"grad_norm": 0.4198344945907593,
"learning_rate": 3.790265684518767e-05,
"loss": 0.3292,
"step": 295
},
{
"epoch": 1.1887550200803212,
"grad_norm": 0.39376020431518555,
"learning_rate": 3.780241013960391e-05,
"loss": 0.3424,
"step": 296
},
{
"epoch": 1.1927710843373494,
"grad_norm": 0.37279409170150757,
"learning_rate": 3.770188363116324e-05,
"loss": 0.261,
"step": 297
},
{
"epoch": 1.1967871485943775,
"grad_norm": 0.44697171449661255,
"learning_rate": 3.760107951692107e-05,
"loss": 0.3212,
"step": 298
},
{
"epoch": 1.2008032128514057,
"grad_norm": 0.38348427414894104,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.2476,
"step": 299
},
{
"epoch": 1.2048192771084336,
"grad_norm": 0.4076741337776184,
"learning_rate": 3.73986472895417e-05,
"loss": 0.3623,
"step": 300
},
{
"epoch": 1.2088353413654618,
"grad_norm": 0.4394407868385315,
"learning_rate": 3.729702360065859e-05,
"loss": 0.3339,
"step": 301
},
{
"epoch": 1.21285140562249,
"grad_norm": 0.39291417598724365,
"learning_rate": 3.7195131154385474e-05,
"loss": 0.289,
"step": 302
},
{
"epoch": 1.216867469879518,
"grad_norm": 0.4359102249145508,
"learning_rate": 3.7092972177631e-05,
"loss": 0.393,
"step": 303
},
{
"epoch": 1.2208835341365463,
"grad_norm": 0.4636111259460449,
"learning_rate": 3.699054890312892e-05,
"loss": 0.3732,
"step": 304
},
{
"epoch": 1.2248995983935742,
"grad_norm": 0.37869930267333984,
"learning_rate": 3.6887863569389386e-05,
"loss": 0.2599,
"step": 305
},
{
"epoch": 1.2289156626506024,
"grad_norm": 0.42484331130981445,
"learning_rate": 3.678491842064995e-05,
"loss": 0.3309,
"step": 306
},
{
"epoch": 1.2329317269076305,
"grad_norm": 0.40696099400520325,
"learning_rate": 3.668171570682655e-05,
"loss": 0.2561,
"step": 307
},
{
"epoch": 1.2369477911646587,
"grad_norm": 0.37397104501724243,
"learning_rate": 3.6578257683464364e-05,
"loss": 0.2373,
"step": 308
},
{
"epoch": 1.2409638554216866,
"grad_norm": 0.41433000564575195,
"learning_rate": 3.6474546611688445e-05,
"loss": 0.35,
"step": 309
},
{
"epoch": 1.2449799196787148,
"grad_norm": 0.4146500825881958,
"learning_rate": 3.6370584758154366e-05,
"loss": 0.3757,
"step": 310
},
{
"epoch": 1.248995983935743,
"grad_norm": 0.41974082589149475,
"learning_rate": 3.6266374394998634e-05,
"loss": 0.3758,
"step": 311
},
{
"epoch": 1.2530120481927711,
"grad_norm": 0.4489721953868866,
"learning_rate": 3.616191779978907e-05,
"loss": 0.3936,
"step": 312
},
{
"epoch": 1.2570281124497993,
"grad_norm": 0.4388364553451538,
"learning_rate": 3.6057217255475034e-05,
"loss": 0.3241,
"step": 313
},
{
"epoch": 1.2610441767068274,
"grad_norm": 0.4363100230693817,
"learning_rate": 3.595227505033747e-05,
"loss": 0.3434,
"step": 314
},
{
"epoch": 1.2650602409638554,
"grad_norm": 0.3850543797016144,
"learning_rate": 3.5847093477938956e-05,
"loss": 0.3019,
"step": 315
},
{
"epoch": 1.2690763052208835,
"grad_norm": 0.4056536555290222,
"learning_rate": 3.574167483707356e-05,
"loss": 0.3634,
"step": 316
},
{
"epoch": 1.2730923694779117,
"grad_norm": 0.37497851252555847,
"learning_rate": 3.5636021431716606e-05,
"loss": 0.3203,
"step": 317
},
{
"epoch": 1.2771084337349397,
"grad_norm": 0.3976558446884155,
"learning_rate": 3.5530135570974274e-05,
"loss": 0.2743,
"step": 318
},
{
"epoch": 1.2811244979919678,
"grad_norm": 0.45948538184165955,
"learning_rate": 3.542401956903321e-05,
"loss": 0.2788,
"step": 319
},
{
"epoch": 1.285140562248996,
"grad_norm": 0.3974132537841797,
"learning_rate": 3.5317675745109866e-05,
"loss": 0.3103,
"step": 320
},
{
"epoch": 1.2891566265060241,
"grad_norm": 0.4223279356956482,
"learning_rate": 3.521110642339991e-05,
"loss": 0.3299,
"step": 321
},
{
"epoch": 1.2931726907630523,
"grad_norm": 0.36853110790252686,
"learning_rate": 3.51043139330273e-05,
"loss": 0.2611,
"step": 322
},
{
"epoch": 1.2971887550200802,
"grad_norm": 0.40147364139556885,
"learning_rate": 3.499730060799352e-05,
"loss": 0.2484,
"step": 323
},
{
"epoch": 1.3012048192771084,
"grad_norm": 0.39763209223747253,
"learning_rate": 3.489006878712647e-05,
"loss": 0.3381,
"step": 324
},
{
"epoch": 1.3052208835341366,
"grad_norm": 0.4128108322620392,
"learning_rate": 3.4782620814029375e-05,
"loss": 0.2844,
"step": 325
},
{
"epoch": 1.3092369477911647,
"grad_norm": 0.4215083420276642,
"learning_rate": 3.4674959037029594e-05,
"loss": 0.288,
"step": 326
},
{
"epoch": 1.3132530120481927,
"grad_norm": 0.45807206630706787,
"learning_rate": 3.456708580912725e-05,
"loss": 0.3709,
"step": 327
},
{
"epoch": 1.3172690763052208,
"grad_norm": 0.38204073905944824,
"learning_rate": 3.4459003487943844e-05,
"loss": 0.2541,
"step": 328
},
{
"epoch": 1.321285140562249,
"grad_norm": 0.42295801639556885,
"learning_rate": 3.4350714435670707e-05,
"loss": 0.3247,
"step": 329
},
{
"epoch": 1.3253012048192772,
"grad_norm": 0.45487505197525024,
"learning_rate": 3.424222101901738e-05,
"loss": 0.3377,
"step": 330
},
{
"epoch": 1.3293172690763053,
"grad_norm": 0.45981574058532715,
"learning_rate": 3.413352560915988e-05,
"loss": 0.3403,
"step": 331
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.3605084717273712,
"learning_rate": 3.4024630581688894e-05,
"loss": 0.242,
"step": 332
},
{
"epoch": 1.3373493975903614,
"grad_norm": 0.4557378888130188,
"learning_rate": 3.391553831655782e-05,
"loss": 0.2929,
"step": 333
},
{
"epoch": 1.3413654618473896,
"grad_norm": 0.5284321904182434,
"learning_rate": 3.380625119803084e-05,
"loss": 0.3434,
"step": 334
},
{
"epoch": 1.3453815261044177,
"grad_norm": 0.4053751230239868,
"learning_rate": 3.369677161463068e-05,
"loss": 0.3149,
"step": 335
},
{
"epoch": 1.3493975903614457,
"grad_norm": 0.4405737519264221,
"learning_rate": 3.358710195908653e-05,
"loss": 0.3668,
"step": 336
},
{
"epoch": 1.3534136546184738,
"grad_norm": 0.3991630971431732,
"learning_rate": 3.3477244628281666e-05,
"loss": 0.3252,
"step": 337
},
{
"epoch": 1.357429718875502,
"grad_norm": 0.39196085929870605,
"learning_rate": 3.336720202320113e-05,
"loss": 0.3282,
"step": 338
},
{
"epoch": 1.3614457831325302,
"grad_norm": 0.38728272914886475,
"learning_rate": 3.3256976548879184e-05,
"loss": 0.2776,
"step": 339
},
{
"epoch": 1.3654618473895583,
"grad_norm": 0.3763287365436554,
"learning_rate": 3.314657061434681e-05,
"loss": 0.2625,
"step": 340
},
{
"epoch": 1.3694779116465863,
"grad_norm": 0.3745321035385132,
"learning_rate": 3.303598663257904e-05,
"loss": 0.2815,
"step": 341
},
{
"epoch": 1.3734939759036144,
"grad_norm": 0.36064666509628296,
"learning_rate": 3.292522702044221e-05,
"loss": 0.2893,
"step": 342
},
{
"epoch": 1.3775100401606426,
"grad_norm": 0.5100922584533691,
"learning_rate": 3.281429419864112e-05,
"loss": 0.3293,
"step": 343
},
{
"epoch": 1.3815261044176708,
"grad_norm": 0.401780903339386,
"learning_rate": 3.270319059166617e-05,
"loss": 0.2773,
"step": 344
},
{
"epoch": 1.3855421686746987,
"grad_norm": 0.4642065465450287,
"learning_rate": 3.259191862774037e-05,
"loss": 0.3379,
"step": 345
},
{
"epoch": 1.3895582329317269,
"grad_norm": 0.39117753505706787,
"learning_rate": 3.248048073876622e-05,
"loss": 0.2766,
"step": 346
},
{
"epoch": 1.393574297188755,
"grad_norm": 0.4060978293418884,
"learning_rate": 3.2368879360272606e-05,
"loss": 0.2647,
"step": 347
},
{
"epoch": 1.3975903614457832,
"grad_norm": 0.4262392520904541,
"learning_rate": 3.225711693136156e-05,
"loss": 0.3177,
"step": 348
},
{
"epoch": 1.4016064257028114,
"grad_norm": 0.41557395458221436,
"learning_rate": 3.214519589465495e-05,
"loss": 0.2425,
"step": 349
},
{
"epoch": 1.4056224899598393,
"grad_norm": 0.4348689019680023,
"learning_rate": 3.203311869624107e-05,
"loss": 0.3529,
"step": 350
},
{
"epoch": 1.4096385542168675,
"grad_norm": 0.4739913046360016,
"learning_rate": 3.1920887785621235e-05,
"loss": 0.3207,
"step": 351
},
{
"epoch": 1.4136546184738956,
"grad_norm": 0.42416346073150635,
"learning_rate": 3.180850561565621e-05,
"loss": 0.2813,
"step": 352
},
{
"epoch": 1.4176706827309236,
"grad_norm": 0.4376218616962433,
"learning_rate": 3.1695974642512585e-05,
"loss": 0.3071,
"step": 353
},
{
"epoch": 1.4216867469879517,
"grad_norm": 0.4232673943042755,
"learning_rate": 3.158329732560912e-05,
"loss": 0.3093,
"step": 354
},
{
"epoch": 1.4257028112449799,
"grad_norm": 0.4880342185497284,
"learning_rate": 3.147047612756302e-05,
"loss": 0.2903,
"step": 355
},
{
"epoch": 1.429718875502008,
"grad_norm": 0.468056857585907,
"learning_rate": 3.1357513514136044e-05,
"loss": 0.3505,
"step": 356
},
{
"epoch": 1.4337349397590362,
"grad_norm": 0.4468645751476288,
"learning_rate": 3.1244411954180676e-05,
"loss": 0.3317,
"step": 357
},
{
"epoch": 1.4377510040160644,
"grad_norm": 0.44746091961860657,
"learning_rate": 3.113117391958612e-05,
"loss": 0.3241,
"step": 358
},
{
"epoch": 1.4417670682730923,
"grad_norm": 0.4584064781665802,
"learning_rate": 3.101780188522433e-05,
"loss": 0.3492,
"step": 359
},
{
"epoch": 1.4457831325301205,
"grad_norm": 0.5009322166442871,
"learning_rate": 3.090429832889586e-05,
"loss": 0.3362,
"step": 360
},
{
"epoch": 1.4497991967871486,
"grad_norm": 0.4156329333782196,
"learning_rate": 3.0790665731275765e-05,
"loss": 0.3024,
"step": 361
},
{
"epoch": 1.4538152610441766,
"grad_norm": 0.44589802622795105,
"learning_rate": 3.0676906575859334e-05,
"loss": 0.2992,
"step": 362
},
{
"epoch": 1.4578313253012047,
"grad_norm": 0.47200101613998413,
"learning_rate": 3.056302334890786e-05,
"loss": 0.3723,
"step": 363
},
{
"epoch": 1.461847389558233,
"grad_norm": 0.37657684087753296,
"learning_rate": 3.0449018539394276e-05,
"loss": 0.2817,
"step": 364
},
{
"epoch": 1.465863453815261,
"grad_norm": 0.4025072753429413,
"learning_rate": 3.0334894638948753e-05,
"loss": 0.3437,
"step": 365
},
{
"epoch": 1.4698795180722892,
"grad_norm": 0.38258296251296997,
"learning_rate": 3.022065414180425e-05,
"loss": 0.3572,
"step": 366
},
{
"epoch": 1.4738955823293174,
"grad_norm": 0.3879828453063965,
"learning_rate": 3.010629954474201e-05,
"loss": 0.2984,
"step": 367
},
{
"epoch": 1.4779116465863453,
"grad_norm": 0.3913291394710541,
"learning_rate": 2.9991833347036985e-05,
"loss": 0.2755,
"step": 368
},
{
"epoch": 1.4819277108433735,
"grad_norm": 0.4224880039691925,
"learning_rate": 2.9877258050403212e-05,
"loss": 0.3324,
"step": 369
},
{
"epoch": 1.4859437751004017,
"grad_norm": 0.4849439859390259,
"learning_rate": 2.9762576158939125e-05,
"loss": 0.3818,
"step": 370
},
{
"epoch": 1.4899598393574296,
"grad_norm": 0.4108419120311737,
"learning_rate": 2.9647790179072872e-05,
"loss": 0.2819,
"step": 371
},
{
"epoch": 1.4939759036144578,
"grad_norm": 0.46975579857826233,
"learning_rate": 2.9532902619507462e-05,
"loss": 0.3213,
"step": 372
},
{
"epoch": 1.497991967871486,
"grad_norm": 0.37872588634490967,
"learning_rate": 2.9417915991166008e-05,
"loss": 0.2886,
"step": 373
},
{
"epoch": 1.502008032128514,
"grad_norm": 0.3882961571216583,
"learning_rate": 2.93028328071368e-05,
"loss": 0.2792,
"step": 374
},
{
"epoch": 1.5060240963855422,
"grad_norm": 0.4001757502555847,
"learning_rate": 2.918765558261841e-05,
"loss": 0.273,
"step": 375
},
{
"epoch": 1.5100401606425704,
"grad_norm": 0.4576755166053772,
"learning_rate": 2.9072386834864724e-05,
"loss": 0.2985,
"step": 376
},
{
"epoch": 1.5140562248995983,
"grad_norm": 0.40884050726890564,
"learning_rate": 2.8957029083129873e-05,
"loss": 0.241,
"step": 377
},
{
"epoch": 1.5180722891566265,
"grad_norm": 0.4215402901172638,
"learning_rate": 2.884158484861325e-05,
"loss": 0.3428,
"step": 378
},
{
"epoch": 1.5220883534136547,
"grad_norm": 0.4656892418861389,
"learning_rate": 2.872605665440436e-05,
"loss": 0.3499,
"step": 379
},
{
"epoch": 1.5261044176706826,
"grad_norm": 0.39561203122138977,
"learning_rate": 2.8610447025427683e-05,
"loss": 0.2815,
"step": 380
},
{
"epoch": 1.5301204819277108,
"grad_norm": 0.4454112648963928,
"learning_rate": 2.849475848838749e-05,
"loss": 0.2744,
"step": 381
},
{
"epoch": 1.534136546184739,
"grad_norm": 0.41716262698173523,
"learning_rate": 2.8378993571712638e-05,
"loss": 0.3507,
"step": 382
},
{
"epoch": 1.538152610441767,
"grad_norm": 0.414767324924469,
"learning_rate": 2.8263154805501297e-05,
"loss": 0.3325,
"step": 383
},
{
"epoch": 1.5421686746987953,
"grad_norm": 0.3738459348678589,
"learning_rate": 2.8147244721465636e-05,
"loss": 0.2559,
"step": 384
},
{
"epoch": 1.5461847389558234,
"grad_norm": 0.45450860261917114,
"learning_rate": 2.8031265852876537e-05,
"loss": 0.3471,
"step": 385
},
{
"epoch": 1.5502008032128514,
"grad_norm": 0.46035537123680115,
"learning_rate": 2.791522073450819e-05,
"loss": 0.3543,
"step": 386
},
{
"epoch": 1.5542168674698795,
"grad_norm": 0.41047337651252747,
"learning_rate": 2.7799111902582696e-05,
"loss": 0.3134,
"step": 387
},
{
"epoch": 1.5582329317269075,
"grad_norm": 0.4487716853618622,
"learning_rate": 2.7682941894714666e-05,
"loss": 0.3284,
"step": 388
},
{
"epoch": 1.5622489959839356,
"grad_norm": 0.511817216873169,
"learning_rate": 2.7566713249855712e-05,
"loss": 0.3324,
"step": 389
},
{
"epoch": 1.5662650602409638,
"grad_norm": 0.4241081774234772,
"learning_rate": 2.7450428508239024e-05,
"loss": 0.31,
"step": 390
},
{
"epoch": 1.570281124497992,
"grad_norm": 0.40205150842666626,
"learning_rate": 2.7334090211323766e-05,
"loss": 0.3453,
"step": 391
},
{
"epoch": 1.5742971887550201,
"grad_norm": 0.44074681401252747,
"learning_rate": 2.7217700901739636e-05,
"loss": 0.3429,
"step": 392
},
{
"epoch": 1.5783132530120483,
"grad_norm": 0.41657623648643494,
"learning_rate": 2.710126312323119e-05,
"loss": 0.3515,
"step": 393
},
{
"epoch": 1.5823293172690764,
"grad_norm": 0.45420968532562256,
"learning_rate": 2.6984779420602323e-05,
"loss": 0.3553,
"step": 394
},
{
"epoch": 1.5863453815261044,
"grad_norm": 0.39352598786354065,
"learning_rate": 2.686825233966061e-05,
"loss": 0.2547,
"step": 395
},
{
"epoch": 1.5903614457831325,
"grad_norm": 0.41205790638923645,
"learning_rate": 2.6751684427161683e-05,
"loss": 0.3538,
"step": 396
},
{
"epoch": 1.5943775100401605,
"grad_norm": 0.4126872718334198,
"learning_rate": 2.663507823075358e-05,
"loss": 0.2893,
"step": 397
},
{
"epoch": 1.5983935742971886,
"grad_norm": 0.4216347932815552,
"learning_rate": 2.651843629892103e-05,
"loss": 0.3279,
"step": 398
},
{
"epoch": 1.6024096385542168,
"grad_norm": 0.3910914361476898,
"learning_rate": 2.6401761180929797e-05,
"loss": 0.2769,
"step": 399
},
{
"epoch": 1.606425702811245,
"grad_norm": 0.4210715591907501,
"learning_rate": 2.6285055426770934e-05,
"loss": 0.302,
"step": 400
},
{
"epoch": 1.6104417670682731,
"grad_norm": 0.4108389914035797,
"learning_rate": 2.616832158710506e-05,
"loss": 0.3253,
"step": 401
},
{
"epoch": 1.6144578313253013,
"grad_norm": 0.41170188784599304,
"learning_rate": 2.6051562213206632e-05,
"loss": 0.2772,
"step": 402
},
{
"epoch": 1.6184738955823295,
"grad_norm": 0.48949119448661804,
"learning_rate": 2.593477985690815e-05,
"loss": 0.388,
"step": 403
},
{
"epoch": 1.6224899598393574,
"grad_norm": 0.44700804352760315,
"learning_rate": 2.5817977070544407e-05,
"loss": 0.3304,
"step": 404
},
{
"epoch": 1.6265060240963856,
"grad_norm": 0.4795211851596832,
"learning_rate": 2.5701156406896725e-05,
"loss": 0.3688,
"step": 405
},
{
"epoch": 1.6305220883534135,
"grad_norm": 0.47693318128585815,
"learning_rate": 2.5584320419137127e-05,
"loss": 0.2896,
"step": 406
},
{
"epoch": 1.6345381526104417,
"grad_norm": 0.5043147802352905,
"learning_rate": 2.5467471660772557e-05,
"loss": 0.3529,
"step": 407
},
{
"epoch": 1.6385542168674698,
"grad_norm": 0.41933324933052063,
"learning_rate": 2.535061268558906e-05,
"loss": 0.3361,
"step": 408
},
{
"epoch": 1.642570281124498,
"grad_norm": 0.41302600502967834,
"learning_rate": 2.5233746047595985e-05,
"loss": 0.2896,
"step": 409
},
{
"epoch": 1.6465863453815262,
"grad_norm": 0.4476950764656067,
"learning_rate": 2.5116874300970138e-05,
"loss": 0.312,
"step": 410
},
{
"epoch": 1.6506024096385543,
"grad_norm": 0.46941718459129333,
"learning_rate": 2.5e-05,
"loss": 0.3004,
"step": 411
},
{
"epoch": 1.6546184738955825,
"grad_norm": 0.44369807839393616,
"learning_rate": 2.4883125699029865e-05,
"loss": 0.3247,
"step": 412
},
{
"epoch": 1.6586345381526104,
"grad_norm": 0.5012811422348022,
"learning_rate": 2.4766253952404024e-05,
"loss": 0.3212,
"step": 413
},
{
"epoch": 1.6626506024096386,
"grad_norm": 0.5274367332458496,
"learning_rate": 2.4649387314410945e-05,
"loss": 0.3282,
"step": 414
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.46215999126434326,
"learning_rate": 2.4532528339227452e-05,
"loss": 0.3086,
"step": 415
},
{
"epoch": 1.6706827309236947,
"grad_norm": 0.45172807574272156,
"learning_rate": 2.441567958086288e-05,
"loss": 0.295,
"step": 416
},
{
"epoch": 1.6746987951807228,
"grad_norm": 0.46284404397010803,
"learning_rate": 2.429884359310328e-05,
"loss": 0.3272,
"step": 417
},
{
"epoch": 1.678714859437751,
"grad_norm": 0.47019344568252563,
"learning_rate": 2.41820229294556e-05,
"loss": 0.2892,
"step": 418
},
{
"epoch": 1.6827309236947792,
"grad_norm": 0.5012633800506592,
"learning_rate": 2.4065220143091864e-05,
"loss": 0.3693,
"step": 419
},
{
"epoch": 1.6867469879518073,
"grad_norm": 0.43759238719940186,
"learning_rate": 2.3948437786793377e-05,
"loss": 0.312,
"step": 420
},
{
"epoch": 1.6907630522088355,
"grad_norm": 0.39541277289390564,
"learning_rate": 2.383167841289495e-05,
"loss": 0.248,
"step": 421
},
{
"epoch": 1.6947791164658634,
"grad_norm": 0.45473945140838623,
"learning_rate": 2.371494457322907e-05,
"loss": 0.2982,
"step": 422
},
{
"epoch": 1.6987951807228916,
"grad_norm": 0.41014575958251953,
"learning_rate": 2.3598238819070202e-05,
"loss": 0.2836,
"step": 423
},
{
"epoch": 1.7028112449799195,
"grad_norm": 0.46837180852890015,
"learning_rate": 2.3481563701078966e-05,
"loss": 0.3009,
"step": 424
},
{
"epoch": 1.7068273092369477,
"grad_norm": 0.41497769951820374,
"learning_rate": 2.3364921769246423e-05,
"loss": 0.3228,
"step": 425
},
{
"epoch": 1.7108433734939759,
"grad_norm": 0.48312485218048096,
"learning_rate": 2.3248315572838316e-05,
"loss": 0.3026,
"step": 426
},
{
"epoch": 1.714859437751004,
"grad_norm": 0.4200379550457001,
"learning_rate": 2.3131747660339394e-05,
"loss": 0.2751,
"step": 427
},
{
"epoch": 1.7188755020080322,
"grad_norm": 0.4631725549697876,
"learning_rate": 2.3015220579397682e-05,
"loss": 0.3425,
"step": 428
},
{
"epoch": 1.7228915662650603,
"grad_norm": 0.46534058451652527,
"learning_rate": 2.2898736876768815e-05,
"loss": 0.3226,
"step": 429
},
{
"epoch": 1.7269076305220885,
"grad_norm": 0.3950774371623993,
"learning_rate": 2.278229909826037e-05,
"loss": 0.2662,
"step": 430
},
{
"epoch": 1.7309236947791165,
"grad_norm": 0.4251087009906769,
"learning_rate": 2.2665909788676237e-05,
"loss": 0.3072,
"step": 431
},
{
"epoch": 1.7349397590361446,
"grad_norm": 0.41468513011932373,
"learning_rate": 2.2549571491760986e-05,
"loss": 0.3047,
"step": 432
},
{
"epoch": 1.7389558232931726,
"grad_norm": 0.4196534752845764,
"learning_rate": 2.2433286750144293e-05,
"loss": 0.2708,
"step": 433
},
{
"epoch": 1.7429718875502007,
"grad_norm": 0.45039716362953186,
"learning_rate": 2.231705810528534e-05,
"loss": 0.3028,
"step": 434
},
{
"epoch": 1.7469879518072289,
"grad_norm": 0.5077593326568604,
"learning_rate": 2.2200888097417307e-05,
"loss": 0.2942,
"step": 435
},
{
"epoch": 1.751004016064257,
"grad_norm": 0.4380891025066376,
"learning_rate": 2.2084779265491813e-05,
"loss": 0.3037,
"step": 436
},
{
"epoch": 1.7550200803212852,
"grad_norm": 0.5840532183647156,
"learning_rate": 2.1968734147123465e-05,
"loss": 0.2832,
"step": 437
},
{
"epoch": 1.7590361445783134,
"grad_norm": 0.4980887472629547,
"learning_rate": 2.185275527853437e-05,
"loss": 0.3245,
"step": 438
},
{
"epoch": 1.7630522088353415,
"grad_norm": 0.4570533037185669,
"learning_rate": 2.173684519449872e-05,
"loss": 0.2917,
"step": 439
},
{
"epoch": 1.7670682730923695,
"grad_norm": 0.45963743329048157,
"learning_rate": 2.162100642828737e-05,
"loss": 0.282,
"step": 440
},
{
"epoch": 1.7710843373493976,
"grad_norm": 0.4363883137702942,
"learning_rate": 2.1505241511612522e-05,
"loss": 0.2665,
"step": 441
},
{
"epoch": 1.7751004016064256,
"grad_norm": 0.5356572866439819,
"learning_rate": 2.138955297457233e-05,
"loss": 0.3612,
"step": 442
},
{
"epoch": 1.7791164658634537,
"grad_norm": 0.505367636680603,
"learning_rate": 2.1273943345595637e-05,
"loss": 0.3098,
"step": 443
},
{
"epoch": 1.783132530120482,
"grad_norm": 0.46621936559677124,
"learning_rate": 2.1158415151386744e-05,
"loss": 0.2991,
"step": 444
},
{
"epoch": 1.78714859437751,
"grad_norm": 0.44183534383773804,
"learning_rate": 2.104297091687013e-05,
"loss": 0.2628,
"step": 445
},
{
"epoch": 1.7911646586345382,
"grad_norm": 0.510328471660614,
"learning_rate": 2.0927613165135285e-05,
"loss": 0.3159,
"step": 446
},
{
"epoch": 1.7951807228915664,
"grad_norm": 0.5044543743133545,
"learning_rate": 2.0812344417381595e-05,
"loss": 0.3069,
"step": 447
},
{
"epoch": 1.7991967871485943,
"grad_norm": 0.4802643656730652,
"learning_rate": 2.0697167192863204e-05,
"loss": 0.2962,
"step": 448
},
{
"epoch": 1.8032128514056225,
"grad_norm": 0.46106359362602234,
"learning_rate": 2.0582084008834e-05,
"loss": 0.2765,
"step": 449
},
{
"epoch": 1.8072289156626506,
"grad_norm": 0.4502885043621063,
"learning_rate": 2.0467097380492544e-05,
"loss": 0.2747,
"step": 450
},
{
"epoch": 1.8112449799196786,
"grad_norm": 0.47659075260162354,
"learning_rate": 2.0352209820927137e-05,
"loss": 0.2936,
"step": 451
},
{
"epoch": 1.8152610441767068,
"grad_norm": 0.444698303937912,
"learning_rate": 2.0237423841060877e-05,
"loss": 0.265,
"step": 452
},
{
"epoch": 1.819277108433735,
"grad_norm": 0.4394044280052185,
"learning_rate": 2.0122741949596797e-05,
"loss": 0.2712,
"step": 453
},
{
"epoch": 1.823293172690763,
"grad_norm": 0.4385037124156952,
"learning_rate": 2.000816665296302e-05,
"loss": 0.3121,
"step": 454
},
{
"epoch": 1.8273092369477912,
"grad_norm": 0.48680782318115234,
"learning_rate": 1.9893700455257996e-05,
"loss": 0.3196,
"step": 455
},
{
"epoch": 1.8313253012048194,
"grad_norm": 0.41616690158843994,
"learning_rate": 1.977934585819576e-05,
"loss": 0.2999,
"step": 456
},
{
"epoch": 1.8353413654618473,
"grad_norm": 0.44171151518821716,
"learning_rate": 1.9665105361051256e-05,
"loss": 0.2971,
"step": 457
},
{
"epoch": 1.8393574297188755,
"grad_norm": 0.42723822593688965,
"learning_rate": 1.9550981460605733e-05,
"loss": 0.2629,
"step": 458
},
{
"epoch": 1.8433734939759037,
"grad_norm": 0.43502771854400635,
"learning_rate": 1.9436976651092144e-05,
"loss": 0.3089,
"step": 459
},
{
"epoch": 1.8473895582329316,
"grad_norm": 0.42654067277908325,
"learning_rate": 1.932309342414067e-05,
"loss": 0.3228,
"step": 460
},
{
"epoch": 1.8514056224899598,
"grad_norm": 0.4032960534095764,
"learning_rate": 1.9209334268724247e-05,
"loss": 0.2395,
"step": 461
},
{
"epoch": 1.855421686746988,
"grad_norm": 0.4426659345626831,
"learning_rate": 1.909570167110415e-05,
"loss": 0.3048,
"step": 462
},
{
"epoch": 1.859437751004016,
"grad_norm": 0.4877111613750458,
"learning_rate": 1.8982198114775682e-05,
"loss": 0.3411,
"step": 463
},
{
"epoch": 1.8634538152610443,
"grad_norm": 0.44326019287109375,
"learning_rate": 1.8868826080413875e-05,
"loss": 0.3135,
"step": 464
},
{
"epoch": 1.8674698795180724,
"grad_norm": 0.500212550163269,
"learning_rate": 1.8755588045819327e-05,
"loss": 0.3221,
"step": 465
},
{
"epoch": 1.8714859437751004,
"grad_norm": 0.43086370825767517,
"learning_rate": 1.864248648586395e-05,
"loss": 0.2936,
"step": 466
},
{
"epoch": 1.8755020080321285,
"grad_norm": 0.42855504155158997,
"learning_rate": 1.852952387243698e-05,
"loss": 0.2468,
"step": 467
},
{
"epoch": 1.8795180722891565,
"grad_norm": 0.5011900663375854,
"learning_rate": 1.841670267439088e-05,
"loss": 0.3038,
"step": 468
},
{
"epoch": 1.8835341365461846,
"grad_norm": 0.44242334365844727,
"learning_rate": 1.8304025357487424e-05,
"loss": 0.3242,
"step": 469
},
{
"epoch": 1.8875502008032128,
"grad_norm": 0.4678678512573242,
"learning_rate": 1.81914943843438e-05,
"loss": 0.307,
"step": 470
},
{
"epoch": 1.891566265060241,
"grad_norm": 0.47704440355300903,
"learning_rate": 1.8079112214378768e-05,
"loss": 0.3092,
"step": 471
},
{
"epoch": 1.895582329317269,
"grad_norm": 0.459638386964798,
"learning_rate": 1.7966881303758938e-05,
"loss": 0.2603,
"step": 472
},
{
"epoch": 1.8995983935742973,
"grad_norm": 0.5107981562614441,
"learning_rate": 1.7854804105345062e-05,
"loss": 0.2853,
"step": 473
},
{
"epoch": 1.9036144578313254,
"grad_norm": 0.5379002690315247,
"learning_rate": 1.7742883068638447e-05,
"loss": 0.3111,
"step": 474
},
{
"epoch": 1.9076305220883534,
"grad_norm": 0.46416106820106506,
"learning_rate": 1.7631120639727393e-05,
"loss": 0.3034,
"step": 475
},
{
"epoch": 1.9116465863453815,
"grad_norm": 0.45781102776527405,
"learning_rate": 1.7519519261233786e-05,
"loss": 0.343,
"step": 476
},
{
"epoch": 1.9156626506024095,
"grad_norm": 0.4929346442222595,
"learning_rate": 1.7408081372259632e-05,
"loss": 0.3325,
"step": 477
},
{
"epoch": 1.9196787148594376,
"grad_norm": 0.4723629355430603,
"learning_rate": 1.729680940833383e-05,
"loss": 0.3035,
"step": 478
},
{
"epoch": 1.9236947791164658,
"grad_norm": 0.42473867535591125,
"learning_rate": 1.7185705801358892e-05,
"loss": 0.2729,
"step": 479
},
{
"epoch": 1.927710843373494,
"grad_norm": 0.4923630356788635,
"learning_rate": 1.7074772979557802e-05,
"loss": 0.3262,
"step": 480
},
{
"epoch": 1.9317269076305221,
"grad_norm": 0.4716297686100006,
"learning_rate": 1.6964013367420966e-05,
"loss": 0.2826,
"step": 481
},
{
"epoch": 1.9357429718875503,
"grad_norm": 0.8560923933982849,
"learning_rate": 1.6853429385653196e-05,
"loss": 0.3091,
"step": 482
},
{
"epoch": 1.9397590361445785,
"grad_norm": 0.43498608469963074,
"learning_rate": 1.6743023451120832e-05,
"loss": 0.3182,
"step": 483
},
{
"epoch": 1.9437751004016064,
"grad_norm": 0.49866142868995667,
"learning_rate": 1.6632797976798886e-05,
"loss": 0.269,
"step": 484
},
{
"epoch": 1.9477911646586346,
"grad_norm": 0.4267672598361969,
"learning_rate": 1.6522755371718333e-05,
"loss": 0.2574,
"step": 485
},
{
"epoch": 1.9518072289156625,
"grad_norm": 0.5241855978965759,
"learning_rate": 1.641289804091347e-05,
"loss": 0.2999,
"step": 486
},
{
"epoch": 1.9558232931726907,
"grad_norm": 0.4990847110748291,
"learning_rate": 1.6303228385369318e-05,
"loss": 0.3732,
"step": 487
},
{
"epoch": 1.9598393574297188,
"grad_norm": 0.5224765539169312,
"learning_rate": 1.6193748801969163e-05,
"loss": 0.3151,
"step": 488
},
{
"epoch": 1.963855421686747,
"grad_norm": 0.45409926772117615,
"learning_rate": 1.6084461683442176e-05,
"loss": 0.2995,
"step": 489
},
{
"epoch": 1.9678714859437751,
"grad_norm": 0.4588085412979126,
"learning_rate": 1.5975369418311112e-05,
"loss": 0.2966,
"step": 490
},
{
"epoch": 1.9718875502008033,
"grad_norm": 0.48878952860832214,
"learning_rate": 1.5866474390840125e-05,
"loss": 0.375,
"step": 491
},
{
"epoch": 1.9759036144578315,
"grad_norm": 0.4290201663970947,
"learning_rate": 1.5757778980982626e-05,
"loss": 0.2723,
"step": 492
},
{
"epoch": 1.9799196787148594,
"grad_norm": 0.44579726457595825,
"learning_rate": 1.5649285564329296e-05,
"loss": 0.2881,
"step": 493
},
{
"epoch": 1.9839357429718876,
"grad_norm": 0.5025163888931274,
"learning_rate": 1.5540996512056165e-05,
"loss": 0.3208,
"step": 494
},
{
"epoch": 1.9879518072289155,
"grad_norm": 0.4830820560455322,
"learning_rate": 1.5432914190872757e-05,
"loss": 0.3005,
"step": 495
},
{
"epoch": 1.9919678714859437,
"grad_norm": 0.47153279185295105,
"learning_rate": 1.5325040962970415e-05,
"loss": 0.3625,
"step": 496
},
{
"epoch": 1.9959839357429718,
"grad_norm": 0.45995303988456726,
"learning_rate": 1.5217379185970629e-05,
"loss": 0.3192,
"step": 497
},
{
"epoch": 2.0,
"grad_norm": 0.4606611132621765,
"learning_rate": 1.5109931212873534e-05,
"loss": 0.2748,
"step": 498
},
{
"epoch": 2.004016064257028,
"grad_norm": 0.49655020236968994,
"learning_rate": 1.500269939200648e-05,
"loss": 0.3096,
"step": 499
},
{
"epoch": 2.0080321285140563,
"grad_norm": 0.40395551919937134,
"learning_rate": 1.4895686066972703e-05,
"loss": 0.2647,
"step": 500
},
{
"epoch": 2.0120481927710845,
"grad_norm": 0.42507630586624146,
"learning_rate": 1.4788893576600099e-05,
"loss": 0.2819,
"step": 501
},
{
"epoch": 2.0160642570281126,
"grad_norm": 0.45520922541618347,
"learning_rate": 1.4682324254890134e-05,
"loss": 0.2894,
"step": 502
},
{
"epoch": 2.0200803212851404,
"grad_norm": 0.4556037485599518,
"learning_rate": 1.4575980430966807e-05,
"loss": 0.279,
"step": 503
},
{
"epoch": 2.0240963855421685,
"grad_norm": 0.45103222131729126,
"learning_rate": 1.446986442902574e-05,
"loss": 0.266,
"step": 504
},
{
"epoch": 2.0281124497991967,
"grad_norm": 0.40927568078041077,
"learning_rate": 1.4363978568283412e-05,
"loss": 0.2397,
"step": 505
},
{
"epoch": 2.032128514056225,
"grad_norm": 0.3951800465583801,
"learning_rate": 1.4258325162926442e-05,
"loss": 0.2009,
"step": 506
},
{
"epoch": 2.036144578313253,
"grad_norm": 0.4924771189689636,
"learning_rate": 1.4152906522061048e-05,
"loss": 0.2887,
"step": 507
},
{
"epoch": 2.040160642570281,
"grad_norm": 0.4882904887199402,
"learning_rate": 1.4047724949662539e-05,
"loss": 0.3229,
"step": 508
},
{
"epoch": 2.0441767068273093,
"grad_norm": 0.46903589367866516,
"learning_rate": 1.3942782744524973e-05,
"loss": 0.2923,
"step": 509
},
{
"epoch": 2.0481927710843375,
"grad_norm": 0.4528999626636505,
"learning_rate": 1.3838082200210931e-05,
"loss": 0.2635,
"step": 510
},
{
"epoch": 2.0522088353413657,
"grad_norm": 0.4789353609085083,
"learning_rate": 1.3733625605001365e-05,
"loss": 0.2102,
"step": 511
},
{
"epoch": 2.0562248995983934,
"grad_norm": 0.4338991045951843,
"learning_rate": 1.362941524184564e-05,
"loss": 0.2442,
"step": 512
},
{
"epoch": 2.0602409638554215,
"grad_norm": 0.4939529001712799,
"learning_rate": 1.3525453388311554e-05,
"loss": 0.2402,
"step": 513
},
{
"epoch": 2.0642570281124497,
"grad_norm": 0.5466839671134949,
"learning_rate": 1.3421742316535637e-05,
"loss": 0.2739,
"step": 514
},
{
"epoch": 2.068273092369478,
"grad_norm": 0.49066290259361267,
"learning_rate": 1.331828429317345e-05,
"loss": 0.2425,
"step": 515
},
{
"epoch": 2.072289156626506,
"grad_norm": 0.5615804195404053,
"learning_rate": 1.3215081579350058e-05,
"loss": 0.2389,
"step": 516
},
{
"epoch": 2.076305220883534,
"grad_norm": 0.49414533376693726,
"learning_rate": 1.3112136430610621e-05,
"loss": 0.2726,
"step": 517
},
{
"epoch": 2.0803212851405624,
"grad_norm": 0.5196560025215149,
"learning_rate": 1.3009451096871083e-05,
"loss": 0.2239,
"step": 518
},
{
"epoch": 2.0843373493975905,
"grad_norm": 0.6148899793624878,
"learning_rate": 1.2907027822369005e-05,
"loss": 0.2568,
"step": 519
},
{
"epoch": 2.0883534136546187,
"grad_norm": 0.5107576847076416,
"learning_rate": 1.2804868845614525e-05,
"loss": 0.2844,
"step": 520
},
{
"epoch": 2.0923694779116464,
"grad_norm": 0.5158917903900146,
"learning_rate": 1.2702976399341421e-05,
"loss": 0.2127,
"step": 521
},
{
"epoch": 2.0963855421686746,
"grad_norm": 0.5524556040763855,
"learning_rate": 1.2601352710458313e-05,
"loss": 0.2409,
"step": 522
},
{
"epoch": 2.1004016064257027,
"grad_norm": 0.5443823933601379,
"learning_rate": 1.2500000000000006e-05,
"loss": 0.2256,
"step": 523
},
{
"epoch": 2.104417670682731,
"grad_norm": 0.5326666831970215,
"learning_rate": 1.2398920483078938e-05,
"loss": 0.2359,
"step": 524
},
{
"epoch": 2.108433734939759,
"grad_norm": 0.5946417450904846,
"learning_rate": 1.229811636883677e-05,
"loss": 0.2744,
"step": 525
},
{
"epoch": 2.112449799196787,
"grad_norm": 0.6217758655548096,
"learning_rate": 1.2197589860396108e-05,
"loss": 0.2345,
"step": 526
},
{
"epoch": 2.1164658634538154,
"grad_norm": 0.5596932172775269,
"learning_rate": 1.2097343154812332e-05,
"loss": 0.2229,
"step": 527
},
{
"epoch": 2.1204819277108435,
"grad_norm": 0.6184597015380859,
"learning_rate": 1.1997378443025634e-05,
"loss": 0.2441,
"step": 528
},
{
"epoch": 2.1244979919678713,
"grad_norm": 0.6836581230163574,
"learning_rate": 1.1897697909813057e-05,
"loss": 0.2922,
"step": 529
},
{
"epoch": 2.1285140562248994,
"grad_norm": 0.6201472282409668,
"learning_rate": 1.1798303733740802e-05,
"loss": 0.2408,
"step": 530
},
{
"epoch": 2.1325301204819276,
"grad_norm": 0.628838837146759,
"learning_rate": 1.1699198087116589e-05,
"loss": 0.2398,
"step": 531
},
{
"epoch": 2.1365461847389557,
"grad_norm": 0.6514667868614197,
"learning_rate": 1.1600383135942191e-05,
"loss": 0.2488,
"step": 532
},
{
"epoch": 2.140562248995984,
"grad_norm": 0.6185716390609741,
"learning_rate": 1.1501861039866094e-05,
"loss": 0.2466,
"step": 533
},
{
"epoch": 2.144578313253012,
"grad_norm": 0.7880752086639404,
"learning_rate": 1.140363395213629e-05,
"loss": 0.2946,
"step": 534
},
{
"epoch": 2.1485943775100402,
"grad_norm": 0.624570906162262,
"learning_rate": 1.130570401955322e-05,
"loss": 0.269,
"step": 535
},
{
"epoch": 2.1526104417670684,
"grad_norm": 0.6464323997497559,
"learning_rate": 1.1208073382422865e-05,
"loss": 0.2976,
"step": 536
},
{
"epoch": 2.1566265060240966,
"grad_norm": 0.5748754143714905,
"learning_rate": 1.1110744174509952e-05,
"loss": 0.264,
"step": 537
},
{
"epoch": 2.1606425702811247,
"grad_norm": 0.5655550360679626,
"learning_rate": 1.1013718522991315e-05,
"loss": 0.2751,
"step": 538
},
{
"epoch": 2.1646586345381524,
"grad_norm": 0.5602949261665344,
"learning_rate": 1.0916998548409449e-05,
"loss": 0.2144,
"step": 539
},
{
"epoch": 2.1686746987951806,
"grad_norm": 0.5831141471862793,
"learning_rate": 1.0820586364626104e-05,
"loss": 0.1952,
"step": 540
},
{
"epoch": 2.1726907630522088,
"grad_norm": 0.5606767535209656,
"learning_rate": 1.072448407877612e-05,
"loss": 0.2365,
"step": 541
},
{
"epoch": 2.176706827309237,
"grad_norm": 0.5832523703575134,
"learning_rate": 1.0628693791221372e-05,
"loss": 0.2627,
"step": 542
},
{
"epoch": 2.180722891566265,
"grad_norm": 0.6273181438446045,
"learning_rate": 1.0533217595504858e-05,
"loss": 0.2724,
"step": 543
},
{
"epoch": 2.1847389558232932,
"grad_norm": 0.5562799572944641,
"learning_rate": 1.043805757830495e-05,
"loss": 0.2364,
"step": 544
},
{
"epoch": 2.1887550200803214,
"grad_norm": 0.5575546622276306,
"learning_rate": 1.0343215819389782e-05,
"loss": 0.277,
"step": 545
},
{
"epoch": 2.1927710843373496,
"grad_norm": 0.6664512157440186,
"learning_rate": 1.0248694391571801e-05,
"loss": 0.2737,
"step": 546
},
{
"epoch": 2.1967871485943773,
"grad_norm": 0.6023972034454346,
"learning_rate": 1.0154495360662464e-05,
"loss": 0.2656,
"step": 547
},
{
"epoch": 2.2008032128514055,
"grad_norm": 0.6038061380386353,
"learning_rate": 1.0060620785427084e-05,
"loss": 0.2493,
"step": 548
},
{
"epoch": 2.2048192771084336,
"grad_norm": 0.5747958421707153,
"learning_rate": 9.967072717539851e-06,
"loss": 0.2445,
"step": 549
},
{
"epoch": 2.208835341365462,
"grad_norm": 1.0454212427139282,
"learning_rate": 9.873853201538971e-06,
"loss": 0.281,
"step": 550
},
{
"epoch": 2.21285140562249,
"grad_norm": 0.629789412021637,
"learning_rate": 9.780964274781984e-06,
"loss": 0.2587,
"step": 551
},
{
"epoch": 2.216867469879518,
"grad_norm": 0.6804355382919312,
"learning_rate": 9.688407967401248e-06,
"loss": 0.2559,
"step": 552
},
{
"epoch": 2.2208835341365463,
"grad_norm": 0.6152710914611816,
"learning_rate": 9.596186302259563e-06,
"loss": 0.2786,
"step": 553
},
{
"epoch": 2.2248995983935744,
"grad_norm": 0.5551489591598511,
"learning_rate": 9.504301294905965e-06,
"loss": 0.2202,
"step": 554
},
{
"epoch": 2.2289156626506026,
"grad_norm": 0.5943081378936768,
"learning_rate": 9.412754953531663e-06,
"loss": 0.2442,
"step": 555
},
{
"epoch": 2.2329317269076308,
"grad_norm": 0.5956800580024719,
"learning_rate": 9.32154927892617e-06,
"loss": 0.2537,
"step": 556
},
{
"epoch": 2.2369477911646585,
"grad_norm": 0.5956325531005859,
"learning_rate": 9.230686264433547e-06,
"loss": 0.2186,
"step": 557
},
{
"epoch": 2.2409638554216866,
"grad_norm": 0.5610763430595398,
"learning_rate": 9.140167895908867e-06,
"loss": 0.2419,
"step": 558
},
{
"epoch": 2.244979919678715,
"grad_norm": 0.5321049094200134,
"learning_rate": 9.049996151674789e-06,
"loss": 0.2042,
"step": 559
},
{
"epoch": 2.248995983935743,
"grad_norm": 0.6597780585289001,
"learning_rate": 8.960173002478336e-06,
"loss": 0.3188,
"step": 560
},
{
"epoch": 2.253012048192771,
"grad_norm": 0.5808085203170776,
"learning_rate": 8.870700411447816e-06,
"loss": 0.2898,
"step": 561
},
{
"epoch": 2.2570281124497993,
"grad_norm": 0.6633182764053345,
"learning_rate": 8.781580334049919e-06,
"loss": 0.2918,
"step": 562
},
{
"epoch": 2.2610441767068274,
"grad_norm": 0.7098970413208008,
"learning_rate": 8.69281471804698e-06,
"loss": 0.2798,
"step": 563
},
{
"epoch": 2.2650602409638556,
"grad_norm": 0.5442720055580139,
"learning_rate": 8.604405503454399e-06,
"loss": 0.2158,
"step": 564
},
{
"epoch": 2.2690763052208833,
"grad_norm": 0.6439300179481506,
"learning_rate": 8.51635462249828e-06,
"loss": 0.2957,
"step": 565
},
{
"epoch": 2.2730923694779115,
"grad_norm": 0.6382263898849487,
"learning_rate": 8.428663999573142e-06,
"loss": 0.3017,
"step": 566
},
{
"epoch": 2.2771084337349397,
"grad_norm": 0.6080713272094727,
"learning_rate": 8.341335551199902e-06,
"loss": 0.2303,
"step": 567
},
{
"epoch": 2.281124497991968,
"grad_norm": 0.5899115800857544,
"learning_rate": 8.25437118598398e-06,
"loss": 0.2673,
"step": 568
},
{
"epoch": 2.285140562248996,
"grad_norm": 0.6265336871147156,
"learning_rate": 8.167772804573571e-06,
"loss": 0.2719,
"step": 569
},
{
"epoch": 2.289156626506024,
"grad_norm": 0.5498580932617188,
"learning_rate": 8.081542299618139e-06,
"loss": 0.2116,
"step": 570
},
{
"epoch": 2.2931726907630523,
"grad_norm": 0.5346493124961853,
"learning_rate": 7.99568155572701e-06,
"loss": 0.2195,
"step": 571
},
{
"epoch": 2.2971887550200805,
"grad_norm": 0.6264692544937134,
"learning_rate": 7.910192449428217e-06,
"loss": 0.2796,
"step": 572
},
{
"epoch": 2.3012048192771086,
"grad_norm": 0.6243094801902771,
"learning_rate": 7.825076849127458e-06,
"loss": 0.2539,
"step": 573
},
{
"epoch": 2.305220883534137,
"grad_norm": 0.6067578792572021,
"learning_rate": 7.740336615067292e-06,
"loss": 0.2764,
"step": 574
},
{
"epoch": 2.3092369477911645,
"grad_norm": 0.575790524482727,
"learning_rate": 7.65597359928646e-06,
"loss": 0.2159,
"step": 575
},
{
"epoch": 2.3132530120481927,
"grad_norm": 0.6838176846504211,
"learning_rate": 7.571989645579419e-06,
"loss": 0.2939,
"step": 576
},
{
"epoch": 2.317269076305221,
"grad_norm": 0.6233852505683899,
"learning_rate": 7.488386589456042e-06,
"loss": 0.2585,
"step": 577
},
{
"epoch": 2.321285140562249,
"grad_norm": 0.8048217296600342,
"learning_rate": 7.405166258101495e-06,
"loss": 0.2752,
"step": 578
},
{
"epoch": 2.325301204819277,
"grad_norm": 0.7208977341651917,
"learning_rate": 7.3223304703363135e-06,
"loss": 0.2896,
"step": 579
},
{
"epoch": 2.3293172690763053,
"grad_norm": 0.6144018173217773,
"learning_rate": 7.239881036576651e-06,
"loss": 0.2668,
"step": 580
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.5961222648620605,
"learning_rate": 7.157819758794698e-06,
"loss": 0.2206,
"step": 581
},
{
"epoch": 2.337349397590361,
"grad_norm": 0.6311051249504089,
"learning_rate": 7.076148430479321e-06,
"loss": 0.2424,
"step": 582
},
{
"epoch": 2.3413654618473894,
"grad_norm": 0.6178223490715027,
"learning_rate": 6.99486883659684e-06,
"loss": 0.2117,
"step": 583
},
{
"epoch": 2.3453815261044175,
"grad_norm": 0.6484620571136475,
"learning_rate": 6.91398275355204e-06,
"loss": 0.2426,
"step": 584
},
{
"epoch": 2.3493975903614457,
"grad_norm": 0.6223502159118652,
"learning_rate": 6.833491949149329e-06,
"loss": 0.2549,
"step": 585
},
{
"epoch": 2.353413654618474,
"grad_norm": 0.644263505935669,
"learning_rate": 6.753398182554116e-06,
"loss": 0.2806,
"step": 586
},
{
"epoch": 2.357429718875502,
"grad_norm": 0.7159098982810974,
"learning_rate": 6.673703204254347e-06,
"loss": 0.2466,
"step": 587
},
{
"epoch": 2.36144578313253,
"grad_norm": 0.6419246196746826,
"learning_rate": 6.594408756022272e-06,
"loss": 0.2309,
"step": 588
},
{
"epoch": 2.3654618473895583,
"grad_norm": 0.7051436305046082,
"learning_rate": 6.515516570876351e-06,
"loss": 0.3082,
"step": 589
},
{
"epoch": 2.3694779116465865,
"grad_norm": 0.6455468535423279,
"learning_rate": 6.437028373043386e-06,
"loss": 0.2162,
"step": 590
},
{
"epoch": 2.3734939759036147,
"grad_norm": 0.6076522469520569,
"learning_rate": 6.358945877920861e-06,
"loss": 0.2462,
"step": 591
},
{
"epoch": 2.3775100401606424,
"grad_norm": 0.7366360425949097,
"learning_rate": 6.281270792039418e-06,
"loss": 0.2718,
"step": 592
},
{
"epoch": 2.3815261044176705,
"grad_norm": 0.680263340473175,
"learning_rate": 6.204004813025568e-06,
"loss": 0.2963,
"step": 593
},
{
"epoch": 2.3855421686746987,
"grad_norm": 0.6568954586982727,
"learning_rate": 6.127149629564605e-06,
"loss": 0.2642,
"step": 594
},
{
"epoch": 2.389558232931727,
"grad_norm": 0.6699411869049072,
"learning_rate": 6.050706921363672e-06,
"loss": 0.2642,
"step": 595
},
{
"epoch": 2.393574297188755,
"grad_norm": 0.6432385444641113,
"learning_rate": 5.974678359115094e-06,
"loss": 0.262,
"step": 596
},
{
"epoch": 2.397590361445783,
"grad_norm": 0.8128492832183838,
"learning_rate": 5.899065604459814e-06,
"loss": 0.2658,
"step": 597
},
{
"epoch": 2.4016064257028114,
"grad_norm": 0.6553678512573242,
"learning_rate": 5.823870309951096e-06,
"loss": 0.2296,
"step": 598
},
{
"epoch": 2.4056224899598395,
"grad_norm": 0.5871982574462891,
"learning_rate": 5.749094119018431e-06,
"loss": 0.2462,
"step": 599
},
{
"epoch": 2.4096385542168672,
"grad_norm": 0.6238443851470947,
"learning_rate": 5.674738665931575e-06,
"loss": 0.1927,
"step": 600
},
{
"epoch": 2.4136546184738954,
"grad_norm": 0.584065318107605,
"learning_rate": 5.600805575764878e-06,
"loss": 0.2203,
"step": 601
},
{
"epoch": 2.4176706827309236,
"grad_norm": 0.6560770273208618,
"learning_rate": 5.527296464361726e-06,
"loss": 0.261,
"step": 602
},
{
"epoch": 2.4216867469879517,
"grad_norm": 0.6104132533073425,
"learning_rate": 5.454212938299255e-06,
"loss": 0.2478,
"step": 603
},
{
"epoch": 2.42570281124498,
"grad_norm": 0.7333542108535767,
"learning_rate": 5.381556594853227e-06,
"loss": 0.2838,
"step": 604
},
{
"epoch": 2.429718875502008,
"grad_norm": 0.6325370073318481,
"learning_rate": 5.3093290219631165e-06,
"loss": 0.2521,
"step": 605
},
{
"epoch": 2.433734939759036,
"grad_norm": 0.7057302594184875,
"learning_rate": 5.237531798197415e-06,
"loss": 0.3002,
"step": 606
},
{
"epoch": 2.4377510040160644,
"grad_norm": 0.6641374826431274,
"learning_rate": 5.166166492719124e-06,
"loss": 0.258,
"step": 607
},
{
"epoch": 2.4417670682730925,
"grad_norm": 0.6088208556175232,
"learning_rate": 5.0952346652514635e-06,
"loss": 0.2192,
"step": 608
},
{
"epoch": 2.4457831325301207,
"grad_norm": 0.5935956835746765,
"learning_rate": 5.02473786604378e-06,
"loss": 0.2297,
"step": 609
},
{
"epoch": 2.4497991967871484,
"grad_norm": 0.6421974897384644,
"learning_rate": 4.954677635837668e-06,
"loss": 0.2639,
"step": 610
},
{
"epoch": 2.4538152610441766,
"grad_norm": 0.795393168926239,
"learning_rate": 4.885055505833291e-06,
"loss": 0.3451,
"step": 611
},
{
"epoch": 2.4578313253012047,
"grad_norm": 0.6571994423866272,
"learning_rate": 4.81587299765594e-06,
"loss": 0.2476,
"step": 612
},
{
"epoch": 2.461847389558233,
"grad_norm": 0.7034909725189209,
"learning_rate": 4.747131623322737e-06,
"loss": 0.296,
"step": 613
},
{
"epoch": 2.465863453815261,
"grad_norm": 0.651319682598114,
"learning_rate": 4.678832885209622e-06,
"loss": 0.269,
"step": 614
},
{
"epoch": 2.4698795180722892,
"grad_norm": 0.7266422510147095,
"learning_rate": 4.610978276018496e-06,
"loss": 0.297,
"step": 615
},
{
"epoch": 2.4738955823293174,
"grad_norm": 0.6697298884391785,
"learning_rate": 4.543569278744625e-06,
"loss": 0.2596,
"step": 616
},
{
"epoch": 2.4779116465863456,
"grad_norm": 0.6902493238449097,
"learning_rate": 4.476607366644192e-06,
"loss": 0.27,
"step": 617
},
{
"epoch": 2.4819277108433733,
"grad_norm": 0.6635175347328186,
"learning_rate": 4.410094003202134e-06,
"loss": 0.2494,
"step": 618
},
{
"epoch": 2.4859437751004014,
"grad_norm": 0.7108765840530396,
"learning_rate": 4.344030642100133e-06,
"loss": 0.2465,
"step": 619
},
{
"epoch": 2.4899598393574296,
"grad_norm": 0.666045069694519,
"learning_rate": 4.278418727184849e-06,
"loss": 0.2253,
"step": 620
},
{
"epoch": 2.4939759036144578,
"grad_norm": 0.631812572479248,
"learning_rate": 4.213259692436367e-06,
"loss": 0.2602,
"step": 621
},
{
"epoch": 2.497991967871486,
"grad_norm": 0.7290167808532715,
"learning_rate": 4.148554961936876e-06,
"loss": 0.2219,
"step": 622
},
{
"epoch": 2.502008032128514,
"grad_norm": 0.6213678121566772,
"learning_rate": 4.0843059498395065e-06,
"loss": 0.2277,
"step": 623
},
{
"epoch": 2.5060240963855422,
"grad_norm": 0.6184942126274109,
"learning_rate": 4.0205140603374465e-06,
"loss": 0.2394,
"step": 624
},
{
"epoch": 2.5100401606425704,
"grad_norm": 0.6213740110397339,
"learning_rate": 3.957180687633252e-06,
"loss": 0.2667,
"step": 625
},
{
"epoch": 2.5140562248995986,
"grad_norm": 0.6418631076812744,
"learning_rate": 3.894307215908371e-06,
"loss": 0.2502,
"step": 626
},
{
"epoch": 2.5180722891566267,
"grad_norm": 0.7205413579940796,
"learning_rate": 3.831895019292897e-06,
"loss": 0.2585,
"step": 627
},
{
"epoch": 2.522088353413655,
"grad_norm": 0.7143741846084595,
"learning_rate": 3.769945461835531e-06,
"loss": 0.2777,
"step": 628
},
{
"epoch": 2.5261044176706826,
"grad_norm": 0.6922659873962402,
"learning_rate": 3.708459897473768e-06,
"loss": 0.2628,
"step": 629
},
{
"epoch": 2.5301204819277108,
"grad_norm": 0.7213859558105469,
"learning_rate": 3.6474396700043158e-06,
"loss": 0.2897,
"step": 630
},
{
"epoch": 2.534136546184739,
"grad_norm": 0.6911467909812927,
"learning_rate": 3.586886113053717e-06,
"loss": 0.2527,
"step": 631
},
{
"epoch": 2.538152610441767,
"grad_norm": 0.6976168155670166,
"learning_rate": 3.5268005500492003e-06,
"loss": 0.3156,
"step": 632
},
{
"epoch": 2.5421686746987953,
"grad_norm": 0.7184569239616394,
"learning_rate": 3.4671842941897765e-06,
"loss": 0.2964,
"step": 633
},
{
"epoch": 2.5461847389558234,
"grad_norm": 0.624638557434082,
"learning_rate": 3.4080386484175024e-06,
"loss": 0.2498,
"step": 634
},
{
"epoch": 2.550200803212851,
"grad_norm": 0.6823080778121948,
"learning_rate": 3.3493649053890326e-06,
"loss": 0.25,
"step": 635
},
{
"epoch": 2.5542168674698793,
"grad_norm": 0.6704211235046387,
"learning_rate": 3.2911643474473646e-06,
"loss": 0.2507,
"step": 636
},
{
"epoch": 2.5582329317269075,
"grad_norm": 0.6562371253967285,
"learning_rate": 3.233438246593795e-06,
"loss": 0.2479,
"step": 637
},
{
"epoch": 2.5622489959839356,
"grad_norm": 0.6533315181732178,
"learning_rate": 3.1761878644601424e-06,
"loss": 0.2201,
"step": 638
},
{
"epoch": 2.566265060240964,
"grad_norm": 0.637791633605957,
"learning_rate": 3.119414452281158e-06,
"loss": 0.2614,
"step": 639
},
{
"epoch": 2.570281124497992,
"grad_norm": 0.6804649233818054,
"learning_rate": 3.0631192508671857e-06,
"loss": 0.2516,
"step": 640
},
{
"epoch": 2.57429718875502,
"grad_norm": 0.6425567865371704,
"learning_rate": 3.007303490577043e-06,
"loss": 0.2459,
"step": 641
},
{
"epoch": 2.5783132530120483,
"grad_norm": 0.6421180963516235,
"learning_rate": 2.9519683912911266e-06,
"loss": 0.2255,
"step": 642
},
{
"epoch": 2.5823293172690764,
"grad_norm": 0.6302490830421448,
"learning_rate": 2.8971151623847587e-06,
"loss": 0.2262,
"step": 643
},
{
"epoch": 2.5863453815261046,
"grad_norm": 0.6398366689682007,
"learning_rate": 2.8427450027017493e-06,
"loss": 0.2311,
"step": 644
},
{
"epoch": 2.5903614457831328,
"grad_norm": 0.7180038690567017,
"learning_rate": 2.788859100528196e-06,
"loss": 0.2564,
"step": 645
},
{
"epoch": 2.5943775100401605,
"grad_norm": 0.7396331429481506,
"learning_rate": 2.7354586335665207e-06,
"loss": 0.239,
"step": 646
},
{
"epoch": 2.5983935742971886,
"grad_norm": 0.6397483944892883,
"learning_rate": 2.682544768909717e-06,
"loss": 0.2229,
"step": 647
},
{
"epoch": 2.602409638554217,
"grad_norm": 0.6396783590316772,
"learning_rate": 2.6301186630158485e-06,
"loss": 0.2266,
"step": 648
},
{
"epoch": 2.606425702811245,
"grad_norm": 0.7305960059165955,
"learning_rate": 2.578181461682794e-06,
"loss": 0.2823,
"step": 649
},
{
"epoch": 2.610441767068273,
"grad_norm": 0.6242882013320923,
"learning_rate": 2.526734300023162e-06,
"loss": 0.2323,
"step": 650
},
{
"epoch": 2.6144578313253013,
"grad_norm": 0.6772400140762329,
"learning_rate": 2.475778302439524e-06,
"loss": 0.2389,
"step": 651
},
{
"epoch": 2.6184738955823295,
"grad_norm": 0.6879493594169617,
"learning_rate": 2.4253145825998137e-06,
"loss": 0.2203,
"step": 652
},
{
"epoch": 2.622489959839357,
"grad_norm": 0.6160506010055542,
"learning_rate": 2.3753442434129997e-06,
"loss": 0.2079,
"step": 653
},
{
"epoch": 2.6265060240963853,
"grad_norm": 0.7234558463096619,
"learning_rate": 2.325868377004986e-06,
"loss": 0.2359,
"step": 654
},
{
"epoch": 2.6305220883534135,
"grad_norm": 0.6611524224281311,
"learning_rate": 2.2768880646947268e-06,
"loss": 0.208,
"step": 655
},
{
"epoch": 2.6345381526104417,
"grad_norm": 0.8209447264671326,
"learning_rate": 2.2284043769706027e-06,
"loss": 0.2654,
"step": 656
},
{
"epoch": 2.63855421686747,
"grad_norm": 0.7226045727729797,
"learning_rate": 2.1804183734670277e-06,
"loss": 0.2587,
"step": 657
},
{
"epoch": 2.642570281124498,
"grad_norm": 0.6614903807640076,
"learning_rate": 2.1329311029412797e-06,
"loss": 0.2389,
"step": 658
},
{
"epoch": 2.646586345381526,
"grad_norm": 0.6674987077713013,
"learning_rate": 2.0859436032505952e-06,
"loss": 0.2267,
"step": 659
},
{
"epoch": 2.6506024096385543,
"grad_norm": 0.700235903263092,
"learning_rate": 2.039456901329473e-06,
"loss": 0.2494,
"step": 660
},
{
"epoch": 2.6546184738955825,
"grad_norm": 0.6528082489967346,
"learning_rate": 1.9934720131672357e-06,
"loss": 0.2357,
"step": 661
},
{
"epoch": 2.6586345381526106,
"grad_norm": 0.617306113243103,
"learning_rate": 1.9479899437858245e-06,
"loss": 0.2233,
"step": 662
},
{
"epoch": 2.662650602409639,
"grad_norm": 0.7248734831809998,
"learning_rate": 1.9030116872178316e-06,
"loss": 0.2546,
"step": 663
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.6827856302261353,
"learning_rate": 1.8585382264847794e-06,
"loss": 0.2382,
"step": 664
},
{
"epoch": 2.6706827309236947,
"grad_norm": 0.7521485686302185,
"learning_rate": 1.8145705335756297e-06,
"loss": 0.3006,
"step": 665
},
{
"epoch": 2.674698795180723,
"grad_norm": 0.6774089932441711,
"learning_rate": 1.771109569425547e-06,
"loss": 0.2137,
"step": 666
},
{
"epoch": 2.678714859437751,
"grad_norm": 0.6956301331520081,
"learning_rate": 1.7281562838948966e-06,
"loss": 0.2474,
"step": 667
},
{
"epoch": 2.682730923694779,
"grad_norm": 0.6630454063415527,
"learning_rate": 1.6857116157484753e-06,
"loss": 0.2242,
"step": 668
},
{
"epoch": 2.6867469879518073,
"grad_norm": 0.7135544419288635,
"learning_rate": 1.6437764926350074e-06,
"loss": 0.2614,
"step": 669
},
{
"epoch": 2.6907630522088355,
"grad_norm": 0.6465809941291809,
"learning_rate": 1.6023518310668618e-06,
"loss": 0.2401,
"step": 670
},
{
"epoch": 2.694779116465863,
"grad_norm": 0.6323262453079224,
"learning_rate": 1.5614385364000228e-06,
"loss": 0.1934,
"step": 671
},
{
"epoch": 2.6987951807228914,
"grad_norm": 0.7006880640983582,
"learning_rate": 1.5210375028143097e-06,
"loss": 0.2708,
"step": 672
},
{
"epoch": 2.7028112449799195,
"grad_norm": 0.7461385726928711,
"learning_rate": 1.4811496132938196e-06,
"loss": 0.2672,
"step": 673
},
{
"epoch": 2.7068273092369477,
"grad_norm": 0.6886598467826843,
"learning_rate": 1.4417757396076425e-06,
"loss": 0.2127,
"step": 674
},
{
"epoch": 2.710843373493976,
"grad_norm": 0.6426632404327393,
"learning_rate": 1.4029167422908107e-06,
"loss": 0.2509,
"step": 675
},
{
"epoch": 2.714859437751004,
"grad_norm": 0.6661653518676758,
"learning_rate": 1.364573470625477e-06,
"loss": 0.2368,
"step": 676
},
{
"epoch": 2.718875502008032,
"grad_norm": 0.7759952545166016,
"learning_rate": 1.3267467626223606e-06,
"loss": 0.2715,
"step": 677
},
{
"epoch": 2.7228915662650603,
"grad_norm": 0.66762375831604,
"learning_rate": 1.2894374450024338e-06,
"loss": 0.2611,
"step": 678
},
{
"epoch": 2.7269076305220885,
"grad_norm": 0.7410826086997986,
"learning_rate": 1.25264633317885e-06,
"loss": 0.2354,
"step": 679
},
{
"epoch": 2.7309236947791167,
"grad_norm": 0.7787061929702759,
"learning_rate": 1.216374231239134e-06,
"loss": 0.2498,
"step": 680
},
{
"epoch": 2.734939759036145,
"grad_norm": 0.6603359580039978,
"learning_rate": 1.180621931927592e-06,
"loss": 0.1978,
"step": 681
},
{
"epoch": 2.7389558232931726,
"grad_norm": 0.6814593076705933,
"learning_rate": 1.1453902166279895e-06,
"loss": 0.2485,
"step": 682
},
{
"epoch": 2.7429718875502007,
"grad_norm": 0.6960440874099731,
"learning_rate": 1.1106798553464804e-06,
"loss": 0.2368,
"step": 683
},
{
"epoch": 2.746987951807229,
"grad_norm": 0.5936889052391052,
"learning_rate": 1.0764916066947794e-06,
"loss": 0.1974,
"step": 684
},
{
"epoch": 2.751004016064257,
"grad_norm": 0.7146962881088257,
"learning_rate": 1.042826217873566e-06,
"loss": 0.263,
"step": 685
},
{
"epoch": 2.755020080321285,
"grad_norm": 0.7414341568946838,
"learning_rate": 1.0096844246561794e-06,
"loss": 0.2933,
"step": 686
},
{
"epoch": 2.7590361445783134,
"grad_norm": 0.6891923546791077,
"learning_rate": 9.770669513725128e-07,
"loss": 0.243,
"step": 687
},
{
"epoch": 2.7630522088353415,
"grad_norm": 0.7602993845939636,
"learning_rate": 9.449745108931985e-07,
"loss": 0.3286,
"step": 688
},
{
"epoch": 2.7670682730923692,
"grad_norm": 0.8095133900642395,
"learning_rate": 9.134078046140248e-07,
"loss": 0.2663,
"step": 689
},
{
"epoch": 2.7710843373493974,
"grad_norm": 0.6787154078483582,
"learning_rate": 8.823675224406053e-07,
"loss": 0.232,
"step": 690
},
{
"epoch": 2.7751004016064256,
"grad_norm": 0.6297564506530762,
"learning_rate": 8.51854342773295e-07,
"loss": 0.2056,
"step": 691
},
{
"epoch": 2.7791164658634537,
"grad_norm": 0.7379974722862244,
"learning_rate": 8.218689324923728e-07,
"loss": 0.2627,
"step": 692
},
{
"epoch": 2.783132530120482,
"grad_norm": 0.7707710266113281,
"learning_rate": 7.924119469434665e-07,
"loss": 0.2647,
"step": 693
},
{
"epoch": 2.78714859437751,
"grad_norm": 0.7104035019874573,
"learning_rate": 7.634840299232171e-07,
"loss": 0.2741,
"step": 694
},
{
"epoch": 2.791164658634538,
"grad_norm": 0.6474725604057312,
"learning_rate": 7.350858136652261e-07,
"loss": 0.2433,
"step": 695
},
{
"epoch": 2.7951807228915664,
"grad_norm": 0.6437653303146362,
"learning_rate": 7.072179188262251e-07,
"loss": 0.2423,
"step": 696
},
{
"epoch": 2.7991967871485945,
"grad_norm": 0.6407147645950317,
"learning_rate": 6.798809544725171e-07,
"loss": 0.2247,
"step": 697
},
{
"epoch": 2.8032128514056227,
"grad_norm": 0.6896359324455261,
"learning_rate": 6.530755180666592e-07,
"loss": 0.2454,
"step": 698
},
{
"epoch": 2.807228915662651,
"grad_norm": 0.7184605598449707,
"learning_rate": 6.268021954544096e-07,
"loss": 0.2884,
"step": 699
},
{
"epoch": 2.8112449799196786,
"grad_norm": 0.6693071126937866,
"learning_rate": 6.01061560851926e-07,
"loss": 0.259,
"step": 700
},
{
"epoch": 2.8152610441767068,
"grad_norm": 0.6582674384117126,
"learning_rate": 5.758541768332071e-07,
"loss": 0.242,
"step": 701
},
{
"epoch": 2.819277108433735,
"grad_norm": 0.7540878057479858,
"learning_rate": 5.5118059431781e-07,
"loss": 0.2682,
"step": 702
},
{
"epoch": 2.823293172690763,
"grad_norm": 0.602482259273529,
"learning_rate": 5.270413525587909e-07,
"loss": 0.1741,
"step": 703
},
{
"epoch": 2.8273092369477912,
"grad_norm": 0.7750258445739746,
"learning_rate": 5.03436979130939e-07,
"loss": 0.2819,
"step": 704
},
{
"epoch": 2.8313253012048194,
"grad_norm": 0.7113935947418213,
"learning_rate": 4.803679899192392e-07,
"loss": 0.2615,
"step": 705
},
{
"epoch": 2.835341365461847,
"grad_norm": 0.7681975960731506,
"learning_rate": 4.57834889107589e-07,
"loss": 0.2557,
"step": 706
},
{
"epoch": 2.8393574297188753,
"grad_norm": 0.7121516466140747,
"learning_rate": 4.358381691677932e-07,
"loss": 0.2274,
"step": 707
},
{
"epoch": 2.8433734939759034,
"grad_norm": 0.6453240513801575,
"learning_rate": 4.1437831084878974e-07,
"loss": 0.2274,
"step": 708
},
{
"epoch": 2.8473895582329316,
"grad_norm": 0.6856228709220886,
"learning_rate": 3.9345578316614396e-07,
"loss": 0.2346,
"step": 709
},
{
"epoch": 2.8514056224899598,
"grad_norm": 0.7098076939582825,
"learning_rate": 3.730710433918039e-07,
"loss": 0.2418,
"step": 710
},
{
"epoch": 2.855421686746988,
"grad_norm": 0.7646916508674622,
"learning_rate": 3.5322453704410286e-07,
"loss": 0.2868,
"step": 711
},
{
"epoch": 2.859437751004016,
"grad_norm": 0.7499619126319885,
"learning_rate": 3.339166978780256e-07,
"loss": 0.2814,
"step": 712
},
{
"epoch": 2.8634538152610443,
"grad_norm": 0.6964211463928223,
"learning_rate": 3.1514794787571854e-07,
"loss": 0.2468,
"step": 713
},
{
"epoch": 2.8674698795180724,
"grad_norm": 0.6901770234107971,
"learning_rate": 2.969186972372806e-07,
"loss": 0.2194,
"step": 714
},
{
"epoch": 2.8714859437751006,
"grad_norm": 0.7522644996643066,
"learning_rate": 2.7922934437178695e-07,
"loss": 0.3136,
"step": 715
},
{
"epoch": 2.8755020080321287,
"grad_norm": 0.79253751039505,
"learning_rate": 2.620802758885876e-07,
"loss": 0.2453,
"step": 716
},
{
"epoch": 2.8795180722891565,
"grad_norm": 0.6912564635276794,
"learning_rate": 2.454718665888589e-07,
"loss": 0.2679,
"step": 717
},
{
"epoch": 2.8835341365461846,
"grad_norm": 0.6885668039321899,
"learning_rate": 2.2940447945740962e-07,
"loss": 0.3014,
"step": 718
},
{
"epoch": 2.887550200803213,
"grad_norm": 0.6475167870521545,
"learning_rate": 2.1387846565474045e-07,
"loss": 0.2299,
"step": 719
},
{
"epoch": 2.891566265060241,
"grad_norm": 0.6608119606971741,
"learning_rate": 1.9889416450938337e-07,
"loss": 0.2277,
"step": 720
},
{
"epoch": 2.895582329317269,
"grad_norm": 0.5878753662109375,
"learning_rate": 1.844519035104797e-07,
"loss": 0.1895,
"step": 721
},
{
"epoch": 2.8995983935742973,
"grad_norm": 0.6588134765625,
"learning_rate": 1.7055199830061652e-07,
"loss": 0.2007,
"step": 722
},
{
"epoch": 2.9036144578313254,
"grad_norm": 0.6344255208969116,
"learning_rate": 1.571947526689349e-07,
"loss": 0.2374,
"step": 723
},
{
"epoch": 2.907630522088353,
"grad_norm": 0.7238723039627075,
"learning_rate": 1.4438045854449355e-07,
"loss": 0.2344,
"step": 724
},
{
"epoch": 2.9116465863453813,
"grad_norm": 0.6409328579902649,
"learning_rate": 1.3210939598987392e-07,
"loss": 0.2229,
"step": 725
},
{
"epoch": 2.9156626506024095,
"grad_norm": 0.7896729111671448,
"learning_rate": 1.2038183319507955e-07,
"loss": 0.2841,
"step": 726
},
{
"epoch": 2.9196787148594376,
"grad_norm": 0.7235094904899597,
"learning_rate": 1.0919802647165466e-07,
"loss": 0.2693,
"step": 727
},
{
"epoch": 2.923694779116466,
"grad_norm": 0.6696702241897583,
"learning_rate": 9.855822024710248e-08,
"loss": 0.2273,
"step": 728
},
{
"epoch": 2.927710843373494,
"grad_norm": 0.7405685186386108,
"learning_rate": 8.846264705952289e-08,
"loss": 0.302,
"step": 729
},
{
"epoch": 2.931726907630522,
"grad_norm": 0.724983274936676,
"learning_rate": 7.891152755254428e-08,
"loss": 0.2216,
"step": 730
},
{
"epoch": 2.9357429718875503,
"grad_norm": 0.8437711000442505,
"learning_rate": 6.990507047049676e-08,
"loss": 0.2934,
"step": 731
},
{
"epoch": 2.9397590361445785,
"grad_norm": 0.7912435531616211,
"learning_rate": 6.14434726538493e-08,
"loss": 0.312,
"step": 732
},
{
"epoch": 2.9437751004016066,
"grad_norm": 0.751061201095581,
"learning_rate": 5.352691903491303e-08,
"loss": 0.2831,
"step": 733
},
{
"epoch": 2.9477911646586348,
"grad_norm": 0.6593888401985168,
"learning_rate": 4.615558263378617e-08,
"loss": 0.2222,
"step": 734
},
{
"epoch": 2.9518072289156625,
"grad_norm": 0.6126686334609985,
"learning_rate": 3.9329624554584884e-08,
"loss": 0.2433,
"step": 735
},
{
"epoch": 2.9558232931726907,
"grad_norm": 0.6581147909164429,
"learning_rate": 3.304919398192663e-08,
"loss": 0.2536,
"step": 736
},
{
"epoch": 2.959839357429719,
"grad_norm": 0.7128841280937195,
"learning_rate": 2.731442817764662e-08,
"loss": 0.2296,
"step": 737
},
{
"epoch": 2.963855421686747,
"grad_norm": 0.7488843202590942,
"learning_rate": 2.2125452477828047e-08,
"loss": 0.2649,
"step": 738
},
{
"epoch": 2.967871485943775,
"grad_norm": 0.6567497849464417,
"learning_rate": 1.7482380290034794e-08,
"loss": 0.221,
"step": 739
},
{
"epoch": 2.9718875502008033,
"grad_norm": 0.6747626662254333,
"learning_rate": 1.3385313090857887e-08,
"loss": 0.2748,
"step": 740
},
{
"epoch": 2.9759036144578315,
"grad_norm": 0.6256674528121948,
"learning_rate": 9.834340423678368e-09,
"loss": 0.2222,
"step": 741
},
{
"epoch": 2.979919678714859,
"grad_norm": 0.7819543480873108,
"learning_rate": 6.829539896716086e-09,
"loss": 0.2571,
"step": 742
},
{
"epoch": 2.9839357429718874,
"grad_norm": 0.6107457876205444,
"learning_rate": 4.370977181339386e-09,
"loss": 0.1988,
"step": 743
},
{
"epoch": 2.9879518072289155,
"grad_norm": 0.6908827424049377,
"learning_rate": 2.4587060106245895e-09,
"loss": 0.2316,
"step": 744
},
{
"epoch": 2.9919678714859437,
"grad_norm": 0.6972522735595703,
"learning_rate": 1.0927681781819355e-09,
"loss": 0.2003,
"step": 745
},
{
"epoch": 2.995983935742972,
"grad_norm": 0.596504271030426,
"learning_rate": 2.7319353723964656e-10,
"loss": 0.199,
"step": 746
},
{
"epoch": 3.0,
"grad_norm": 0.6581512093544006,
"learning_rate": 0.0,
"loss": 0.2254,
"step": 747
},
{
"epoch": 3.0,
"step": 747,
"total_flos": 7.676283830535586e+18,
"train_loss": 0.3990471245174906,
"train_runtime": 52416.2628,
"train_samples_per_second": 1.824,
"train_steps_per_second": 0.014
}
],
"logging_steps": 1,
"max_steps": 747,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7.676283830535586e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}