Files
qwen2.5-3b-sft/trainer_state.json
ModelHub XC 41c28fe6c0 初始化项目,由ModelHub XC社区提供模型
Model: boradorish/qwen2.5-3b-sft
Source: Original Platform
2026-10-05 23:55:25 +08:00

5504 lines
133 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 780,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0038554216867469878,
"grad_norm": 0.6997007131576538,
"learning_rate": 0.0,
"loss": 0.0405,
"step": 1
},
{
"epoch": 0.0077108433734939755,
"grad_norm": 0.6004841327667236,
"learning_rate": 5.128205128205128e-07,
"loss": 0.0413,
"step": 2
},
{
"epoch": 0.011566265060240964,
"grad_norm": 0.6796688437461853,
"learning_rate": 1.0256410256410257e-06,
"loss": 0.049,
"step": 3
},
{
"epoch": 0.015421686746987951,
"grad_norm": 0.6812688708305359,
"learning_rate": 1.5384615384615387e-06,
"loss": 0.0486,
"step": 4
},
{
"epoch": 0.01927710843373494,
"grad_norm": 0.6957109570503235,
"learning_rate": 2.0512820512820513e-06,
"loss": 0.033,
"step": 5
},
{
"epoch": 0.02313253012048193,
"grad_norm": 0.6179328560829163,
"learning_rate": 2.564102564102564e-06,
"loss": 0.0505,
"step": 6
},
{
"epoch": 0.026987951807228915,
"grad_norm": 0.5707671046257019,
"learning_rate": 3.0769230769230774e-06,
"loss": 0.0409,
"step": 7
},
{
"epoch": 0.030843373493975902,
"grad_norm": 0.41436561942100525,
"learning_rate": 3.58974358974359e-06,
"loss": 0.0433,
"step": 8
},
{
"epoch": 0.03469879518072289,
"grad_norm": 0.3194092810153961,
"learning_rate": 4.102564102564103e-06,
"loss": 0.032,
"step": 9
},
{
"epoch": 0.03855421686746988,
"grad_norm": 0.3646799325942993,
"learning_rate": 4.615384615384616e-06,
"loss": 0.0326,
"step": 10
},
{
"epoch": 0.042409638554216866,
"grad_norm": 0.3623705804347992,
"learning_rate": 5.128205128205128e-06,
"loss": 0.0415,
"step": 11
},
{
"epoch": 0.04626506024096386,
"grad_norm": 0.23459671437740326,
"learning_rate": 5.641025641025641e-06,
"loss": 0.0356,
"step": 12
},
{
"epoch": 0.05012048192771084,
"grad_norm": 0.27805373072624207,
"learning_rate": 6.153846153846155e-06,
"loss": 0.0265,
"step": 13
},
{
"epoch": 0.05397590361445783,
"grad_norm": 0.2236052304506302,
"learning_rate": 6.666666666666667e-06,
"loss": 0.0237,
"step": 14
},
{
"epoch": 0.05783132530120482,
"grad_norm": 0.2813984453678131,
"learning_rate": 7.17948717948718e-06,
"loss": 0.0271,
"step": 15
},
{
"epoch": 0.061686746987951804,
"grad_norm": 0.24619753658771515,
"learning_rate": 7.692307692307694e-06,
"loss": 0.0237,
"step": 16
},
{
"epoch": 0.0655421686746988,
"grad_norm": 0.2730517089366913,
"learning_rate": 8.205128205128205e-06,
"loss": 0.0227,
"step": 17
},
{
"epoch": 0.06939759036144579,
"grad_norm": 0.2350538820028305,
"learning_rate": 8.717948717948719e-06,
"loss": 0.0245,
"step": 18
},
{
"epoch": 0.07325301204819278,
"grad_norm": 0.18193207681179047,
"learning_rate": 9.230769230769232e-06,
"loss": 0.0168,
"step": 19
},
{
"epoch": 0.07710843373493977,
"grad_norm": 0.18165752291679382,
"learning_rate": 9.743589743589744e-06,
"loss": 0.0159,
"step": 20
},
{
"epoch": 0.08096385542168674,
"grad_norm": 0.27797478437423706,
"learning_rate": 1.0256410256410256e-05,
"loss": 0.0301,
"step": 21
},
{
"epoch": 0.08481927710843373,
"grad_norm": 0.26159757375717163,
"learning_rate": 1.076923076923077e-05,
"loss": 0.0279,
"step": 22
},
{
"epoch": 0.08867469879518072,
"grad_norm": 0.20290671288967133,
"learning_rate": 1.1282051282051283e-05,
"loss": 0.0215,
"step": 23
},
{
"epoch": 0.09253012048192771,
"grad_norm": 0.3016839921474457,
"learning_rate": 1.1794871794871796e-05,
"loss": 0.0239,
"step": 24
},
{
"epoch": 0.0963855421686747,
"grad_norm": 0.23746410012245178,
"learning_rate": 1.230769230769231e-05,
"loss": 0.0287,
"step": 25
},
{
"epoch": 0.10024096385542168,
"grad_norm": 0.20524649322032928,
"learning_rate": 1.2820512820512823e-05,
"loss": 0.0233,
"step": 26
},
{
"epoch": 0.10409638554216867,
"grad_norm": 0.23074588179588318,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.0227,
"step": 27
},
{
"epoch": 0.10795180722891566,
"grad_norm": 0.17793533205986023,
"learning_rate": 1.3846153846153847e-05,
"loss": 0.0206,
"step": 28
},
{
"epoch": 0.11180722891566265,
"grad_norm": 0.19907526671886444,
"learning_rate": 1.435897435897436e-05,
"loss": 0.0291,
"step": 29
},
{
"epoch": 0.11566265060240964,
"grad_norm": 0.22600896656513214,
"learning_rate": 1.4871794871794874e-05,
"loss": 0.0277,
"step": 30
},
{
"epoch": 0.11951807228915663,
"grad_norm": 0.2135571539402008,
"learning_rate": 1.5384615384615387e-05,
"loss": 0.0157,
"step": 31
},
{
"epoch": 0.12337349397590361,
"grad_norm": 0.17844079434871674,
"learning_rate": 1.5897435897435897e-05,
"loss": 0.0207,
"step": 32
},
{
"epoch": 0.1272289156626506,
"grad_norm": 0.21249860525131226,
"learning_rate": 1.641025641025641e-05,
"loss": 0.0258,
"step": 33
},
{
"epoch": 0.1310843373493976,
"grad_norm": 0.18014496564865112,
"learning_rate": 1.6923076923076924e-05,
"loss": 0.0187,
"step": 34
},
{
"epoch": 0.13493975903614458,
"grad_norm": 0.1790035516023636,
"learning_rate": 1.7435897435897438e-05,
"loss": 0.0141,
"step": 35
},
{
"epoch": 0.13879518072289157,
"grad_norm": 0.18319612741470337,
"learning_rate": 1.794871794871795e-05,
"loss": 0.0153,
"step": 36
},
{
"epoch": 0.14265060240963856,
"grad_norm": 0.16664963960647583,
"learning_rate": 1.8461538461538465e-05,
"loss": 0.0146,
"step": 37
},
{
"epoch": 0.14650602409638555,
"grad_norm": 0.17741602659225464,
"learning_rate": 1.8974358974358975e-05,
"loss": 0.0138,
"step": 38
},
{
"epoch": 0.15036144578313254,
"grad_norm": 0.16826853156089783,
"learning_rate": 1.9487179487179488e-05,
"loss": 0.0149,
"step": 39
},
{
"epoch": 0.15421686746987953,
"grad_norm": 0.17632468044757843,
"learning_rate": 2e-05,
"loss": 0.0315,
"step": 40
},
{
"epoch": 0.1580722891566265,
"grad_norm": 0.21154020726680756,
"learning_rate": 2.0512820512820512e-05,
"loss": 0.0313,
"step": 41
},
{
"epoch": 0.16192771084337348,
"grad_norm": 0.17007827758789062,
"learning_rate": 2.102564102564103e-05,
"loss": 0.0283,
"step": 42
},
{
"epoch": 0.16578313253012048,
"grad_norm": 0.20888379216194153,
"learning_rate": 2.153846153846154e-05,
"loss": 0.0173,
"step": 43
},
{
"epoch": 0.16963855421686747,
"grad_norm": 0.17302851378917694,
"learning_rate": 2.2051282051282056e-05,
"loss": 0.023,
"step": 44
},
{
"epoch": 0.17349397590361446,
"grad_norm": 0.28571709990501404,
"learning_rate": 2.2564102564102566e-05,
"loss": 0.0528,
"step": 45
},
{
"epoch": 0.17734939759036145,
"grad_norm": 0.2224675714969635,
"learning_rate": 2.3076923076923076e-05,
"loss": 0.0326,
"step": 46
},
{
"epoch": 0.18120481927710844,
"grad_norm": 0.16989563405513763,
"learning_rate": 2.3589743589743593e-05,
"loss": 0.0106,
"step": 47
},
{
"epoch": 0.18506024096385543,
"grad_norm": 0.17279551923274994,
"learning_rate": 2.4102564102564103e-05,
"loss": 0.0143,
"step": 48
},
{
"epoch": 0.18891566265060242,
"grad_norm": 0.1571756899356842,
"learning_rate": 2.461538461538462e-05,
"loss": 0.0112,
"step": 49
},
{
"epoch": 0.1927710843373494,
"grad_norm": 0.21857869625091553,
"learning_rate": 2.512820512820513e-05,
"loss": 0.0149,
"step": 50
},
{
"epoch": 0.1966265060240964,
"grad_norm": 0.21509476006031036,
"learning_rate": 2.5641025641025646e-05,
"loss": 0.0248,
"step": 51
},
{
"epoch": 0.20048192771084336,
"grad_norm": 0.15649166703224182,
"learning_rate": 2.6153846153846157e-05,
"loss": 0.0101,
"step": 52
},
{
"epoch": 0.20433734939759035,
"grad_norm": 0.17392459511756897,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.0115,
"step": 53
},
{
"epoch": 0.20819277108433734,
"grad_norm": 0.13562685251235962,
"learning_rate": 2.7179487179487183e-05,
"loss": 0.0102,
"step": 54
},
{
"epoch": 0.21204819277108433,
"grad_norm": 0.22402575612068176,
"learning_rate": 2.7692307692307694e-05,
"loss": 0.0164,
"step": 55
},
{
"epoch": 0.21590361445783132,
"grad_norm": 0.1317652314901352,
"learning_rate": 2.820512820512821e-05,
"loss": 0.0091,
"step": 56
},
{
"epoch": 0.2197590361445783,
"grad_norm": 0.11563073843717575,
"learning_rate": 2.871794871794872e-05,
"loss": 0.009,
"step": 57
},
{
"epoch": 0.2236144578313253,
"grad_norm": 0.12244722992181778,
"learning_rate": 2.923076923076923e-05,
"loss": 0.0077,
"step": 58
},
{
"epoch": 0.2274698795180723,
"grad_norm": 0.18595360219478607,
"learning_rate": 2.9743589743589747e-05,
"loss": 0.0118,
"step": 59
},
{
"epoch": 0.23132530120481928,
"grad_norm": 0.16396519541740417,
"learning_rate": 3.0256410256410257e-05,
"loss": 0.0113,
"step": 60
},
{
"epoch": 0.23518072289156627,
"grad_norm": 0.16600239276885986,
"learning_rate": 3.0769230769230774e-05,
"loss": 0.0273,
"step": 61
},
{
"epoch": 0.23903614457831326,
"grad_norm": 0.14368009567260742,
"learning_rate": 3.128205128205129e-05,
"loss": 0.0107,
"step": 62
},
{
"epoch": 0.24289156626506025,
"grad_norm": 0.11151304095983505,
"learning_rate": 3.1794871794871795e-05,
"loss": 0.0114,
"step": 63
},
{
"epoch": 0.24674698795180722,
"grad_norm": 0.16931672394275665,
"learning_rate": 3.230769230769231e-05,
"loss": 0.0101,
"step": 64
},
{
"epoch": 0.25060240963855424,
"grad_norm": 0.13678501546382904,
"learning_rate": 3.282051282051282e-05,
"loss": 0.0074,
"step": 65
},
{
"epoch": 0.2544578313253012,
"grad_norm": 0.12231199443340302,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.0117,
"step": 66
},
{
"epoch": 0.2583132530120482,
"grad_norm": 0.2577515244483948,
"learning_rate": 3.384615384615385e-05,
"loss": 0.0304,
"step": 67
},
{
"epoch": 0.2621686746987952,
"grad_norm": 0.10441906005144119,
"learning_rate": 3.435897435897436e-05,
"loss": 0.0077,
"step": 68
},
{
"epoch": 0.26602409638554214,
"grad_norm": 0.18340735137462616,
"learning_rate": 3.4871794871794875e-05,
"loss": 0.0131,
"step": 69
},
{
"epoch": 0.26987951807228916,
"grad_norm": 0.26318177580833435,
"learning_rate": 3.538461538461539e-05,
"loss": 0.0166,
"step": 70
},
{
"epoch": 0.2737349397590361,
"grad_norm": 0.0868140459060669,
"learning_rate": 3.58974358974359e-05,
"loss": 0.0079,
"step": 71
},
{
"epoch": 0.27759036144578314,
"grad_norm": 0.1882234513759613,
"learning_rate": 3.6410256410256416e-05,
"loss": 0.0113,
"step": 72
},
{
"epoch": 0.2814457831325301,
"grad_norm": 0.1313207447528839,
"learning_rate": 3.692307692307693e-05,
"loss": 0.01,
"step": 73
},
{
"epoch": 0.2853012048192771,
"grad_norm": 0.21643561124801636,
"learning_rate": 3.7435897435897436e-05,
"loss": 0.0126,
"step": 74
},
{
"epoch": 0.2891566265060241,
"grad_norm": 0.35117974877357483,
"learning_rate": 3.794871794871795e-05,
"loss": 0.0322,
"step": 75
},
{
"epoch": 0.2930120481927711,
"grad_norm": 0.14641538262367249,
"learning_rate": 3.846153846153846e-05,
"loss": 0.0086,
"step": 76
},
{
"epoch": 0.29686746987951806,
"grad_norm": 0.18411016464233398,
"learning_rate": 3.8974358974358976e-05,
"loss": 0.0254,
"step": 77
},
{
"epoch": 0.3007228915662651,
"grad_norm": 0.17950740456581116,
"learning_rate": 3.948717948717949e-05,
"loss": 0.0191,
"step": 78
},
{
"epoch": 0.30457831325301205,
"grad_norm": 0.14514976739883423,
"learning_rate": 4e-05,
"loss": 0.0107,
"step": 79
},
{
"epoch": 0.30843373493975906,
"grad_norm": 0.22098544239997864,
"learning_rate": 3.999979972589652e-05,
"loss": 0.0176,
"step": 80
},
{
"epoch": 0.312289156626506,
"grad_norm": 0.11959940195083618,
"learning_rate": 3.9999198907597046e-05,
"loss": 0.0168,
"step": 81
},
{
"epoch": 0.316144578313253,
"grad_norm": 0.13479092717170715,
"learning_rate": 3.999819755713442e-05,
"loss": 0.0246,
"step": 82
},
{
"epoch": 0.32,
"grad_norm": 0.15680080652236938,
"learning_rate": 3.9996795694563096e-05,
"loss": 0.0147,
"step": 83
},
{
"epoch": 0.32385542168674697,
"grad_norm": 0.1106104925274849,
"learning_rate": 3.999499334795875e-05,
"loss": 0.017,
"step": 84
},
{
"epoch": 0.327710843373494,
"grad_norm": 0.13254176080226898,
"learning_rate": 3.999279055341771e-05,
"loss": 0.0096,
"step": 85
},
{
"epoch": 0.33156626506024095,
"grad_norm": 0.1391618400812149,
"learning_rate": 3.999018735505626e-05,
"loss": 0.0165,
"step": 86
},
{
"epoch": 0.33542168674698797,
"grad_norm": 0.15080322325229645,
"learning_rate": 3.998718380500971e-05,
"loss": 0.0127,
"step": 87
},
{
"epoch": 0.33927710843373493,
"grad_norm": 0.20322079956531525,
"learning_rate": 3.998377996343139e-05,
"loss": 0.053,
"step": 88
},
{
"epoch": 0.34313253012048195,
"grad_norm": 0.10287915915250778,
"learning_rate": 3.997997589849145e-05,
"loss": 0.0112,
"step": 89
},
{
"epoch": 0.3469879518072289,
"grad_norm": 0.14580729603767395,
"learning_rate": 3.9975771686375434e-05,
"loss": 0.009,
"step": 90
},
{
"epoch": 0.35084337349397593,
"grad_norm": 0.16750818490982056,
"learning_rate": 3.9971167411282835e-05,
"loss": 0.0137,
"step": 91
},
{
"epoch": 0.3546987951807229,
"grad_norm": 0.0841616690158844,
"learning_rate": 3.996616316542537e-05,
"loss": 0.0054,
"step": 92
},
{
"epoch": 0.35855421686746985,
"grad_norm": 0.20032115280628204,
"learning_rate": 3.99607590490251e-05,
"loss": 0.0138,
"step": 93
},
{
"epoch": 0.3624096385542169,
"grad_norm": 0.20202049612998962,
"learning_rate": 3.9954955170312504e-05,
"loss": 0.0128,
"step": 94
},
{
"epoch": 0.36626506024096384,
"grad_norm": 0.10456931591033936,
"learning_rate": 3.9948751645524235e-05,
"loss": 0.0092,
"step": 95
},
{
"epoch": 0.37012048192771085,
"grad_norm": 0.403626948595047,
"learning_rate": 3.994214859890083e-05,
"loss": 0.0206,
"step": 96
},
{
"epoch": 0.3739759036144578,
"grad_norm": 0.11969339102506638,
"learning_rate": 3.9935146162684206e-05,
"loss": 0.0101,
"step": 97
},
{
"epoch": 0.37783132530120483,
"grad_norm": 0.3118373453617096,
"learning_rate": 3.992774447711503e-05,
"loss": 0.0273,
"step": 98
},
{
"epoch": 0.3816867469879518,
"grad_norm": 0.2649889588356018,
"learning_rate": 3.9919943690429906e-05,
"loss": 0.0386,
"step": 99
},
{
"epoch": 0.3855421686746988,
"grad_norm": 0.12110533565282822,
"learning_rate": 3.9911743958858376e-05,
"loss": 0.0144,
"step": 100
},
{
"epoch": 0.3893975903614458,
"grad_norm": 0.15890805423259735,
"learning_rate": 3.9903145446619837e-05,
"loss": 0.0145,
"step": 101
},
{
"epoch": 0.3932530120481928,
"grad_norm": 0.18432742357254028,
"learning_rate": 3.9894148325920226e-05,
"loss": 0.0193,
"step": 102
},
{
"epoch": 0.39710843373493976,
"grad_norm": 0.10168084502220154,
"learning_rate": 3.9884752776948564e-05,
"loss": 0.0159,
"step": 103
},
{
"epoch": 0.4009638554216867,
"grad_norm": 0.1423511803150177,
"learning_rate": 3.987495898787336e-05,
"loss": 0.0106,
"step": 104
},
{
"epoch": 0.40481927710843374,
"grad_norm": 0.12045685201883316,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.0251,
"step": 105
},
{
"epoch": 0.4086746987951807,
"grad_norm": 0.2883131802082062,
"learning_rate": 3.985417748196108e-05,
"loss": 0.0134,
"step": 106
},
{
"epoch": 0.4125301204819277,
"grad_norm": 0.11933720856904984,
"learning_rate": 3.9843190181323744e-05,
"loss": 0.0128,
"step": 107
},
{
"epoch": 0.4163855421686747,
"grad_norm": 0.13842257857322693,
"learning_rate": 3.983180547297404e-05,
"loss": 0.0177,
"step": 108
},
{
"epoch": 0.4202409638554217,
"grad_norm": 0.10292483121156693,
"learning_rate": 3.982002358491817e-05,
"loss": 0.0145,
"step": 109
},
{
"epoch": 0.42409638554216866,
"grad_norm": 0.13685393333435059,
"learning_rate": 3.9807844753116864e-05,
"loss": 0.0184,
"step": 110
},
{
"epoch": 0.4279518072289157,
"grad_norm": 0.11288538575172424,
"learning_rate": 3.979526922148058e-05,
"loss": 0.0238,
"step": 111
},
{
"epoch": 0.43180722891566264,
"grad_norm": 0.18025726079940796,
"learning_rate": 3.978229724186464e-05,
"loss": 0.0135,
"step": 112
},
{
"epoch": 0.43566265060240966,
"grad_norm": 0.1295086145401001,
"learning_rate": 3.9768929074064206e-05,
"loss": 0.0109,
"step": 113
},
{
"epoch": 0.4395180722891566,
"grad_norm": 0.16575755178928375,
"learning_rate": 3.975516498580907e-05,
"loss": 0.0163,
"step": 114
},
{
"epoch": 0.4433734939759036,
"grad_norm": 0.08723966032266617,
"learning_rate": 3.9741005252758255e-05,
"loss": 0.0098,
"step": 115
},
{
"epoch": 0.4472289156626506,
"grad_norm": 0.29995423555374146,
"learning_rate": 3.972645015849457e-05,
"loss": 0.0279,
"step": 116
},
{
"epoch": 0.45108433734939757,
"grad_norm": 0.12310135364532471,
"learning_rate": 3.971149999451886e-05,
"loss": 0.0276,
"step": 117
},
{
"epoch": 0.4549397590361446,
"grad_norm": 0.22947722673416138,
"learning_rate": 3.9696155060244166e-05,
"loss": 0.0143,
"step": 118
},
{
"epoch": 0.45879518072289155,
"grad_norm": 0.1598968803882599,
"learning_rate": 3.9680415662989806e-05,
"loss": 0.0184,
"step": 119
},
{
"epoch": 0.46265060240963857,
"grad_norm": 0.11719726771116257,
"learning_rate": 3.966428211797514e-05,
"loss": 0.0089,
"step": 120
},
{
"epoch": 0.46650602409638553,
"grad_norm": 0.2086484581232071,
"learning_rate": 3.9647754748313294e-05,
"loss": 0.025,
"step": 121
},
{
"epoch": 0.47036144578313255,
"grad_norm": 0.1487833857536316,
"learning_rate": 3.963083388500469e-05,
"loss": 0.0107,
"step": 122
},
{
"epoch": 0.4742168674698795,
"grad_norm": 0.14869804680347443,
"learning_rate": 3.96135198669304e-05,
"loss": 0.0174,
"step": 123
},
{
"epoch": 0.47807228915662653,
"grad_norm": 0.20308156311511993,
"learning_rate": 3.959581304084536e-05,
"loss": 0.0116,
"step": 124
},
{
"epoch": 0.4819277108433735,
"grad_norm": 0.17671343684196472,
"learning_rate": 3.957771376137144e-05,
"loss": 0.0182,
"step": 125
},
{
"epoch": 0.4857831325301205,
"grad_norm": 0.23302067816257477,
"learning_rate": 3.955922239099035e-05,
"loss": 0.013,
"step": 126
},
{
"epoch": 0.48963855421686747,
"grad_norm": 0.11614499241113663,
"learning_rate": 3.954033930003634e-05,
"loss": 0.0134,
"step": 127
},
{
"epoch": 0.49349397590361443,
"grad_norm": 0.1320069283246994,
"learning_rate": 3.952106486668884e-05,
"loss": 0.0097,
"step": 128
},
{
"epoch": 0.49734939759036145,
"grad_norm": 0.11876890808343887,
"learning_rate": 3.9501399476964806e-05,
"loss": 0.0105,
"step": 129
},
{
"epoch": 0.5012048192771085,
"grad_norm": 0.26870131492614746,
"learning_rate": 3.9481343524711095e-05,
"loss": 0.0157,
"step": 130
},
{
"epoch": 0.5050602409638554,
"grad_norm": 0.12262927740812302,
"learning_rate": 3.946089741159648e-05,
"loss": 0.0089,
"step": 131
},
{
"epoch": 0.5089156626506024,
"grad_norm": 0.16786855459213257,
"learning_rate": 3.944006154710366e-05,
"loss": 0.0149,
"step": 132
},
{
"epoch": 0.5127710843373494,
"grad_norm": 0.17113229632377625,
"learning_rate": 3.9418836348521045e-05,
"loss": 0.0167,
"step": 133
},
{
"epoch": 0.5166265060240964,
"grad_norm": 0.1865008920431137,
"learning_rate": 3.939722224093439e-05,
"loss": 0.0172,
"step": 134
},
{
"epoch": 0.5204819277108433,
"grad_norm": 0.11489243060350418,
"learning_rate": 3.937521965721831e-05,
"loss": 0.0135,
"step": 135
},
{
"epoch": 0.5243373493975904,
"grad_norm": 0.12064780294895172,
"learning_rate": 3.9352829038027564e-05,
"loss": 0.0224,
"step": 136
},
{
"epoch": 0.5281927710843374,
"grad_norm": 0.21562877297401428,
"learning_rate": 3.933005083178828e-05,
"loss": 0.0197,
"step": 137
},
{
"epoch": 0.5320481927710843,
"grad_norm": 0.15050241351127625,
"learning_rate": 3.930688549468894e-05,
"loss": 0.0316,
"step": 138
},
{
"epoch": 0.5359036144578313,
"grad_norm": 0.17046259343624115,
"learning_rate": 3.928333349067125e-05,
"loss": 0.0233,
"step": 139
},
{
"epoch": 0.5397590361445783,
"grad_norm": 0.15832003951072693,
"learning_rate": 3.9259395291420864e-05,
"loss": 0.0131,
"step": 140
},
{
"epoch": 0.5436144578313253,
"grad_norm": 0.10384412109851837,
"learning_rate": 3.923507137635792e-05,
"loss": 0.0084,
"step": 141
},
{
"epoch": 0.5474698795180722,
"grad_norm": 0.1093461811542511,
"learning_rate": 3.921036223262745e-05,
"loss": 0.0082,
"step": 142
},
{
"epoch": 0.5513253012048193,
"grad_norm": 0.26693451404571533,
"learning_rate": 3.9185268355089606e-05,
"loss": 0.0198,
"step": 143
},
{
"epoch": 0.5551807228915663,
"grad_norm": 0.10376594215631485,
"learning_rate": 3.915979024630978e-05,
"loss": 0.011,
"step": 144
},
{
"epoch": 0.5590361445783133,
"grad_norm": 0.18130183219909668,
"learning_rate": 3.913392841654851e-05,
"loss": 0.0111,
"step": 145
},
{
"epoch": 0.5628915662650602,
"grad_norm": 0.1697569340467453,
"learning_rate": 3.9107683383751264e-05,
"loss": 0.0188,
"step": 146
},
{
"epoch": 0.5667469879518072,
"grad_norm": 0.1838744878768921,
"learning_rate": 3.9081055673538093e-05,
"loss": 0.0181,
"step": 147
},
{
"epoch": 0.5706024096385542,
"grad_norm": 0.19388824701309204,
"learning_rate": 3.9054045819193074e-05,
"loss": 0.0136,
"step": 148
},
{
"epoch": 0.5744578313253013,
"grad_norm": 0.14233097434043884,
"learning_rate": 3.902665436165364e-05,
"loss": 0.0285,
"step": 149
},
{
"epoch": 0.5783132530120482,
"grad_norm": 0.14576232433319092,
"learning_rate": 3.8998881849499756e-05,
"loss": 0.0139,
"step": 150
},
{
"epoch": 0.5821686746987952,
"grad_norm": 0.12378726154565811,
"learning_rate": 3.897072883894291e-05,
"loss": 0.0154,
"step": 151
},
{
"epoch": 0.5860240963855422,
"grad_norm": 0.180088609457016,
"learning_rate": 3.894219589381501e-05,
"loss": 0.0284,
"step": 152
},
{
"epoch": 0.5898795180722891,
"grad_norm": 0.1278950572013855,
"learning_rate": 3.8913283585557054e-05,
"loss": 0.006,
"step": 153
},
{
"epoch": 0.5937349397590361,
"grad_norm": 0.09193156659603119,
"learning_rate": 3.8883992493207696e-05,
"loss": 0.0098,
"step": 154
},
{
"epoch": 0.5975903614457831,
"grad_norm": 0.0820469856262207,
"learning_rate": 3.885432320339167e-05,
"loss": 0.008,
"step": 155
},
{
"epoch": 0.6014457831325302,
"grad_norm": 0.1358262449502945,
"learning_rate": 3.8824276310308e-05,
"loss": 0.0134,
"step": 156
},
{
"epoch": 0.6053012048192771,
"grad_norm": 0.1384797841310501,
"learning_rate": 3.879385241571817e-05,
"loss": 0.0211,
"step": 157
},
{
"epoch": 0.6091566265060241,
"grad_norm": 0.1795671582221985,
"learning_rate": 3.876305212893399e-05,
"loss": 0.0189,
"step": 158
},
{
"epoch": 0.6130120481927711,
"grad_norm": 0.09805607050657272,
"learning_rate": 3.873187606680543e-05,
"loss": 0.0074,
"step": 159
},
{
"epoch": 0.6168674698795181,
"grad_norm": 0.10765130817890167,
"learning_rate": 3.8700324853708304e-05,
"loss": 0.0268,
"step": 160
},
{
"epoch": 0.620722891566265,
"grad_norm": 0.15209679305553436,
"learning_rate": 3.866839912153168e-05,
"loss": 0.0148,
"step": 161
},
{
"epoch": 0.624578313253012,
"grad_norm": 0.10442651808261871,
"learning_rate": 3.863609950966531e-05,
"loss": 0.0168,
"step": 162
},
{
"epoch": 0.6284337349397591,
"grad_norm": 0.11361811310052872,
"learning_rate": 3.860342666498677e-05,
"loss": 0.0079,
"step": 163
},
{
"epoch": 0.632289156626506,
"grad_norm": 0.09062547236680984,
"learning_rate": 3.8570381241848534e-05,
"loss": 0.0129,
"step": 164
},
{
"epoch": 0.636144578313253,
"grad_norm": 0.10144251585006714,
"learning_rate": 3.853696390206484e-05,
"loss": 0.0179,
"step": 165
},
{
"epoch": 0.64,
"grad_norm": 0.10997909307479858,
"learning_rate": 3.850317531489848e-05,
"loss": 0.0218,
"step": 166
},
{
"epoch": 0.643855421686747,
"grad_norm": 0.08607764542102814,
"learning_rate": 3.846901615704734e-05,
"loss": 0.0124,
"step": 167
},
{
"epoch": 0.6477108433734939,
"grad_norm": 0.20264504849910736,
"learning_rate": 3.843448711263089e-05,
"loss": 0.0189,
"step": 168
},
{
"epoch": 0.651566265060241,
"grad_norm": 0.10201306641101837,
"learning_rate": 3.839958887317649e-05,
"loss": 0.013,
"step": 169
},
{
"epoch": 0.655421686746988,
"grad_norm": 0.15065032243728638,
"learning_rate": 3.8364322137605484e-05,
"loss": 0.0225,
"step": 170
},
{
"epoch": 0.659277108433735,
"grad_norm": 0.1181226447224617,
"learning_rate": 3.832868761221926e-05,
"loss": 0.0117,
"step": 171
},
{
"epoch": 0.6631325301204819,
"grad_norm": 0.277878075838089,
"learning_rate": 3.829268601068509e-05,
"loss": 0.0117,
"step": 172
},
{
"epoch": 0.6669879518072289,
"grad_norm": 0.18172211945056915,
"learning_rate": 3.825631805402182e-05,
"loss": 0.0195,
"step": 173
},
{
"epoch": 0.6708433734939759,
"grad_norm": 0.21377064287662506,
"learning_rate": 3.821958447058543e-05,
"loss": 0.0151,
"step": 174
},
{
"epoch": 0.6746987951807228,
"grad_norm": 0.14039507508277893,
"learning_rate": 3.818248599605448e-05,
"loss": 0.0138,
"step": 175
},
{
"epoch": 0.6785542168674699,
"grad_norm": 0.15213626623153687,
"learning_rate": 3.8145023373415326e-05,
"loss": 0.013,
"step": 176
},
{
"epoch": 0.6824096385542169,
"grad_norm": 0.2724161446094513,
"learning_rate": 3.810719735294731e-05,
"loss": 0.0301,
"step": 177
},
{
"epoch": 0.6862650602409639,
"grad_norm": 0.13877442479133606,
"learning_rate": 3.806900869220765e-05,
"loss": 0.0157,
"step": 178
},
{
"epoch": 0.6901204819277108,
"grad_norm": 0.0968877524137497,
"learning_rate": 3.8030458156016326e-05,
"loss": 0.012,
"step": 179
},
{
"epoch": 0.6939759036144578,
"grad_norm": 0.1643582135438919,
"learning_rate": 3.799154651644075e-05,
"loss": 0.0155,
"step": 180
},
{
"epoch": 0.6978313253012048,
"grad_norm": 0.13681621849536896,
"learning_rate": 3.795227455278029e-05,
"loss": 0.0085,
"step": 181
},
{
"epoch": 0.7016867469879519,
"grad_norm": 0.11992929130792618,
"learning_rate": 3.7912643051550674e-05,
"loss": 0.011,
"step": 182
},
{
"epoch": 0.7055421686746988,
"grad_norm": 0.16314208507537842,
"learning_rate": 3.787265280646825e-05,
"loss": 0.0197,
"step": 183
},
{
"epoch": 0.7093975903614458,
"grad_norm": 0.08850007504224777,
"learning_rate": 3.783230461843406e-05,
"loss": 0.009,
"step": 184
},
{
"epoch": 0.7132530120481928,
"grad_norm": 0.12203098088502884,
"learning_rate": 3.7791599295517825e-05,
"loss": 0.0091,
"step": 185
},
{
"epoch": 0.7171084337349397,
"grad_norm": 0.12872737646102905,
"learning_rate": 3.775053765294175e-05,
"loss": 0.0101,
"step": 186
},
{
"epoch": 0.7209638554216867,
"grad_norm": 0.114129438996315,
"learning_rate": 3.7709120513064196e-05,
"loss": 0.0077,
"step": 187
},
{
"epoch": 0.7248192771084337,
"grad_norm": 0.12316344678401947,
"learning_rate": 3.7667348705363227e-05,
"loss": 0.0103,
"step": 188
},
{
"epoch": 0.7286746987951808,
"grad_norm": 0.07499723881483078,
"learning_rate": 3.762522306641998e-05,
"loss": 0.0059,
"step": 189
},
{
"epoch": 0.7325301204819277,
"grad_norm": 0.10113950073719025,
"learning_rate": 3.75827444399019e-05,
"loss": 0.0137,
"step": 190
},
{
"epoch": 0.7363855421686747,
"grad_norm": 0.10856342315673828,
"learning_rate": 3.7539913676545874e-05,
"loss": 0.0091,
"step": 191
},
{
"epoch": 0.7402409638554217,
"grad_norm": 0.18475328385829926,
"learning_rate": 3.749673163414117e-05,
"loss": 0.0094,
"step": 192
},
{
"epoch": 0.7440963855421687,
"grad_norm": 0.0707508996129036,
"learning_rate": 3.745319917751229e-05,
"loss": 0.0062,
"step": 193
},
{
"epoch": 0.7479518072289156,
"grad_norm": 0.10805204510688782,
"learning_rate": 3.740931717850159e-05,
"loss": 0.0111,
"step": 194
},
{
"epoch": 0.7518072289156627,
"grad_norm": 0.10874560475349426,
"learning_rate": 3.736508651595188e-05,
"loss": 0.0093,
"step": 195
},
{
"epoch": 0.7556626506024097,
"grad_norm": 0.12469008564949036,
"learning_rate": 3.732050807568878e-05,
"loss": 0.012,
"step": 196
},
{
"epoch": 0.7595180722891566,
"grad_norm": 0.0760217234492302,
"learning_rate": 3.727558275050301e-05,
"loss": 0.0068,
"step": 197
},
{
"epoch": 0.7633734939759036,
"grad_norm": 0.1542130559682846,
"learning_rate": 3.7230311440132494e-05,
"loss": 0.0172,
"step": 198
},
{
"epoch": 0.7672289156626506,
"grad_norm": 0.14483506977558136,
"learning_rate": 3.718469505124434e-05,
"loss": 0.0213,
"step": 199
},
{
"epoch": 0.7710843373493976,
"grad_norm": 0.1366746425628662,
"learning_rate": 3.713873449741669e-05,
"loss": 0.0093,
"step": 200
},
{
"epoch": 0.7749397590361445,
"grad_norm": 0.12670476734638214,
"learning_rate": 3.709243069912041e-05,
"loss": 0.0092,
"step": 201
},
{
"epoch": 0.7787951807228916,
"grad_norm": 0.3366965353488922,
"learning_rate": 3.704578458370067e-05,
"loss": 0.017,
"step": 202
},
{
"epoch": 0.7826506024096386,
"grad_norm": 0.1790863275527954,
"learning_rate": 3.699879708535838e-05,
"loss": 0.0247,
"step": 203
},
{
"epoch": 0.7865060240963856,
"grad_norm": 0.20145761966705322,
"learning_rate": 3.695146914513142e-05,
"loss": 0.0075,
"step": 204
},
{
"epoch": 0.7903614457831325,
"grad_norm": 0.09072735905647278,
"learning_rate": 3.69038017108759e-05,
"loss": 0.013,
"step": 205
},
{
"epoch": 0.7942168674698795,
"grad_norm": 1.0004081726074219,
"learning_rate": 3.685579573724707e-05,
"loss": 0.0135,
"step": 206
},
{
"epoch": 0.7980722891566265,
"grad_norm": 7.910810470581055,
"learning_rate": 3.680745218568026e-05,
"loss": 0.0986,
"step": 207
},
{
"epoch": 0.8019277108433734,
"grad_norm": 0.19183991849422455,
"learning_rate": 3.6758772024371626e-05,
"loss": 0.0112,
"step": 208
},
{
"epoch": 0.8057831325301205,
"grad_norm": 0.17618803679943085,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.023,
"step": 209
},
{
"epoch": 0.8096385542168675,
"grad_norm": 0.17773428559303284,
"learning_rate": 3.6660405779001036e-05,
"loss": 0.0166,
"step": 210
},
{
"epoch": 0.8134939759036145,
"grad_norm": 0.11507993191480637,
"learning_rate": 3.6610721664960236e-05,
"loss": 0.008,
"step": 211
},
{
"epoch": 0.8173493975903614,
"grad_norm": 0.21720387041568756,
"learning_rate": 3.656070488118047e-05,
"loss": 0.0364,
"step": 212
},
{
"epoch": 0.8212048192771084,
"grad_norm": 0.1748373657464981,
"learning_rate": 3.65103564293684e-05,
"loss": 0.0121,
"step": 213
},
{
"epoch": 0.8250602409638554,
"grad_norm": 0.11724656075239182,
"learning_rate": 3.645967731787313e-05,
"loss": 0.0178,
"step": 214
},
{
"epoch": 0.8289156626506025,
"grad_norm": 0.13730894029140472,
"learning_rate": 3.640866856166601e-05,
"loss": 0.0078,
"step": 215
},
{
"epoch": 0.8327710843373494,
"grad_norm": 0.14834603667259216,
"learning_rate": 3.635733118232034e-05,
"loss": 0.0148,
"step": 216
},
{
"epoch": 0.8366265060240964,
"grad_norm": 0.14735879004001617,
"learning_rate": 3.6305666207990886e-05,
"loss": 0.0166,
"step": 217
},
{
"epoch": 0.8404819277108434,
"grad_norm": 0.13057921826839447,
"learning_rate": 3.625367467339329e-05,
"loss": 0.0186,
"step": 218
},
{
"epoch": 0.8443373493975903,
"grad_norm": 0.19306808710098267,
"learning_rate": 3.6201357619783336e-05,
"loss": 0.0152,
"step": 219
},
{
"epoch": 0.8481927710843373,
"grad_norm": 0.1420055627822876,
"learning_rate": 3.614871609493614e-05,
"loss": 0.0131,
"step": 220
},
{
"epoch": 0.8520481927710843,
"grad_norm": 0.11154947429895401,
"learning_rate": 3.609575115312511e-05,
"loss": 0.0103,
"step": 221
},
{
"epoch": 0.8559036144578314,
"grad_norm": 0.07021575421094894,
"learning_rate": 3.604246385510088e-05,
"loss": 0.0054,
"step": 222
},
{
"epoch": 0.8597590361445783,
"grad_norm": 0.13351315259933472,
"learning_rate": 3.598885526807003e-05,
"loss": 0.0141,
"step": 223
},
{
"epoch": 0.8636144578313253,
"grad_norm": 0.43515488505363464,
"learning_rate": 3.593492646567372e-05,
"loss": 0.0174,
"step": 224
},
{
"epoch": 0.8674698795180723,
"grad_norm": 0.14238739013671875,
"learning_rate": 3.5880678527966224e-05,
"loss": 0.013,
"step": 225
},
{
"epoch": 0.8713253012048193,
"grad_norm": 0.11826923489570618,
"learning_rate": 3.582611254139324e-05,
"loss": 0.0149,
"step": 226
},
{
"epoch": 0.8751807228915662,
"grad_norm": 0.22871695458889008,
"learning_rate": 3.577122959877017e-05,
"loss": 0.0394,
"step": 227
},
{
"epoch": 0.8790361445783132,
"grad_norm": 0.20841334760189056,
"learning_rate": 3.571603079926024e-05,
"loss": 0.0163,
"step": 228
},
{
"epoch": 0.8828915662650603,
"grad_norm": 0.1635965257883072,
"learning_rate": 3.566051724835245e-05,
"loss": 0.0103,
"step": 229
},
{
"epoch": 0.8867469879518072,
"grad_norm": 0.11963551491498947,
"learning_rate": 3.5604690057839456e-05,
"loss": 0.0121,
"step": 230
},
{
"epoch": 0.8906024096385542,
"grad_norm": 0.14931157231330872,
"learning_rate": 3.554855034579532e-05,
"loss": 0.0224,
"step": 231
},
{
"epoch": 0.8944578313253012,
"grad_norm": 0.2835037410259247,
"learning_rate": 3.549209923655309e-05,
"loss": 0.0282,
"step": 232
},
{
"epoch": 0.8983132530120482,
"grad_norm": 0.18211285769939423,
"learning_rate": 3.5435337860682304e-05,
"loss": 0.012,
"step": 233
},
{
"epoch": 0.9021686746987951,
"grad_norm": 0.18591780960559845,
"learning_rate": 3.537826735496632e-05,
"loss": 0.015,
"step": 234
},
{
"epoch": 0.9060240963855422,
"grad_norm": 0.09304528683423996,
"learning_rate": 3.532088886237956e-05,
"loss": 0.0087,
"step": 235
},
{
"epoch": 0.9098795180722892,
"grad_norm": 0.1557615101337433,
"learning_rate": 3.5263203532064674e-05,
"loss": 0.018,
"step": 236
},
{
"epoch": 0.9137349397590362,
"grad_norm": 0.14380118250846863,
"learning_rate": 3.520521251930941e-05,
"loss": 0.0106,
"step": 237
},
{
"epoch": 0.9175903614457831,
"grad_norm": 0.31495094299316406,
"learning_rate": 3.5146916985523604e-05,
"loss": 0.0186,
"step": 238
},
{
"epoch": 0.9214457831325301,
"grad_norm": 0.16599956154823303,
"learning_rate": 3.5088318098215805e-05,
"loss": 0.0092,
"step": 239
},
{
"epoch": 0.9253012048192771,
"grad_norm": 0.13166308403015137,
"learning_rate": 3.5029417030970005e-05,
"loss": 0.0111,
"step": 240
},
{
"epoch": 0.929156626506024,
"grad_norm": 0.1272827833890915,
"learning_rate": 3.497021496342203e-05,
"loss": 0.0192,
"step": 241
},
{
"epoch": 0.9330120481927711,
"grad_norm": 0.11934683471918106,
"learning_rate": 3.4910713081235975e-05,
"loss": 0.0099,
"step": 242
},
{
"epoch": 0.9368674698795181,
"grad_norm": 0.2631380259990692,
"learning_rate": 3.485091257608047e-05,
"loss": 0.016,
"step": 243
},
{
"epoch": 0.9407228915662651,
"grad_norm": 0.3094984292984009,
"learning_rate": 3.479081464560475e-05,
"loss": 0.0309,
"step": 244
},
{
"epoch": 0.944578313253012,
"grad_norm": 0.11957540363073349,
"learning_rate": 3.473042049341474e-05,
"loss": 0.0125,
"step": 245
},
{
"epoch": 0.948433734939759,
"grad_norm": 0.08568521589040756,
"learning_rate": 3.466973132904892e-05,
"loss": 0.0086,
"step": 246
},
{
"epoch": 0.952289156626506,
"grad_norm": 0.16930826008319855,
"learning_rate": 3.4608748367954064e-05,
"loss": 0.011,
"step": 247
},
{
"epoch": 0.9561445783132531,
"grad_norm": 0.1169271171092987,
"learning_rate": 3.4547472831460976e-05,
"loss": 0.0085,
"step": 248
},
{
"epoch": 0.96,
"grad_norm": 0.16872183978557587,
"learning_rate": 3.4485905946759965e-05,
"loss": 0.0144,
"step": 249
},
{
"epoch": 0.963855421686747,
"grad_norm": 0.09693744778633118,
"learning_rate": 3.442404894687629e-05,
"loss": 0.006,
"step": 250
},
{
"epoch": 0.967710843373494,
"grad_norm": 0.17783288657665253,
"learning_rate": 3.4361903070645484e-05,
"loss": 0.0124,
"step": 251
},
{
"epoch": 0.971566265060241,
"grad_norm": 0.2039792239665985,
"learning_rate": 3.4299469562688494e-05,
"loss": 0.0218,
"step": 252
},
{
"epoch": 0.9754216867469879,
"grad_norm": 0.12866197526454926,
"learning_rate": 3.423674967338681e-05,
"loss": 0.0186,
"step": 253
},
{
"epoch": 0.9792771084337349,
"grad_norm": 0.19284434616565704,
"learning_rate": 3.4173744658857403e-05,
"loss": 0.0112,
"step": 254
},
{
"epoch": 0.983132530120482,
"grad_norm": 0.16508746147155762,
"learning_rate": 3.411045578092754e-05,
"loss": 0.0187,
"step": 255
},
{
"epoch": 0.9869879518072289,
"grad_norm": 0.12111517041921616,
"learning_rate": 3.404688430710955e-05,
"loss": 0.0143,
"step": 256
},
{
"epoch": 0.9908433734939759,
"grad_norm": 0.13150326907634735,
"learning_rate": 3.398303151057543e-05,
"loss": 0.0107,
"step": 257
},
{
"epoch": 0.9946987951807229,
"grad_norm": 0.12251593172550201,
"learning_rate": 3.391889867013134e-05,
"loss": 0.014,
"step": 258
},
{
"epoch": 0.9985542168674699,
"grad_norm": 0.09811683744192123,
"learning_rate": 3.385448707019199e-05,
"loss": 0.0109,
"step": 259
},
{
"epoch": 1.0,
"grad_norm": 0.229774609208107,
"learning_rate": 3.3789798000754924e-05,
"loss": 0.0096,
"step": 260
},
{
"epoch": 1.003855421686747,
"grad_norm": 0.10416769236326218,
"learning_rate": 3.372483275737468e-05,
"loss": 0.0132,
"step": 261
},
{
"epoch": 1.007710843373494,
"grad_norm": 0.1602071076631546,
"learning_rate": 3.365959264113683e-05,
"loss": 0.0085,
"step": 262
},
{
"epoch": 1.011566265060241,
"grad_norm": 0.1573389619588852,
"learning_rate": 3.359407895863199e-05,
"loss": 0.0273,
"step": 263
},
{
"epoch": 1.0154216867469879,
"grad_norm": 0.0807509571313858,
"learning_rate": 3.3528293021929523e-05,
"loss": 0.0176,
"step": 264
},
{
"epoch": 1.0192771084337349,
"grad_norm": 0.12698227167129517,
"learning_rate": 3.34622361485514e-05,
"loss": 0.0275,
"step": 265
},
{
"epoch": 1.0231325301204819,
"grad_norm": 0.07560165226459503,
"learning_rate": 3.339590966144574e-05,
"loss": 0.003,
"step": 266
},
{
"epoch": 1.026987951807229,
"grad_norm": 0.11409149318933487,
"learning_rate": 3.332931488896029e-05,
"loss": 0.0076,
"step": 267
},
{
"epoch": 1.030843373493976,
"grad_norm": 0.11506525427103043,
"learning_rate": 3.326245316481591e-05,
"loss": 0.0151,
"step": 268
},
{
"epoch": 1.034698795180723,
"grad_norm": 0.09818965941667557,
"learning_rate": 3.319532582807977e-05,
"loss": 0.0135,
"step": 269
},
{
"epoch": 1.03855421686747,
"grad_norm": 0.117642842233181,
"learning_rate": 3.31279342231386e-05,
"loss": 0.0075,
"step": 270
},
{
"epoch": 1.0424096385542168,
"grad_norm": 0.12259496003389359,
"learning_rate": 3.30602796996717e-05,
"loss": 0.0108,
"step": 271
},
{
"epoch": 1.0462650602409638,
"grad_norm": 0.18426232039928436,
"learning_rate": 3.299236361262401e-05,
"loss": 0.0062,
"step": 272
},
{
"epoch": 1.0501204819277108,
"grad_norm": 0.10463278740644455,
"learning_rate": 3.2924187322178865e-05,
"loss": 0.0131,
"step": 273
},
{
"epoch": 1.0539759036144578,
"grad_norm": 0.07446189224720001,
"learning_rate": 3.285575219373079e-05,
"loss": 0.0039,
"step": 274
},
{
"epoch": 1.0578313253012048,
"grad_norm": 0.09327566623687744,
"learning_rate": 3.278705959785821e-05,
"loss": 0.0076,
"step": 275
},
{
"epoch": 1.0616867469879518,
"grad_norm": 0.1121094599366188,
"learning_rate": 3.27181109102959e-05,
"loss": 0.0042,
"step": 276
},
{
"epoch": 1.0655421686746989,
"grad_norm": 0.0797392874956131,
"learning_rate": 3.2648907511907544e-05,
"loss": 0.0055,
"step": 277
},
{
"epoch": 1.0693975903614459,
"grad_norm": 0.1773054301738739,
"learning_rate": 3.2579450788657997e-05,
"loss": 0.0196,
"step": 278
},
{
"epoch": 1.0732530120481927,
"grad_norm": 0.8156651258468628,
"learning_rate": 3.250974213158555e-05,
"loss": 0.0096,
"step": 279
},
{
"epoch": 1.0771084337349397,
"grad_norm": 0.10387098044157028,
"learning_rate": 3.243978293677408e-05,
"loss": 0.0071,
"step": 280
},
{
"epoch": 1.0809638554216867,
"grad_norm": 0.29274797439575195,
"learning_rate": 3.23695746053251e-05,
"loss": 0.0147,
"step": 281
},
{
"epoch": 1.0848192771084337,
"grad_norm": 0.0871700718998909,
"learning_rate": 3.229911854332967e-05,
"loss": 0.0105,
"step": 282
},
{
"epoch": 1.0886746987951808,
"grad_norm": 0.13531728088855743,
"learning_rate": 3.222841616184025e-05,
"loss": 0.0188,
"step": 283
},
{
"epoch": 1.0925301204819278,
"grad_norm": 0.11029268801212311,
"learning_rate": 3.215746887684244e-05,
"loss": 0.0067,
"step": 284
},
{
"epoch": 1.0963855421686748,
"grad_norm": 0.10946464538574219,
"learning_rate": 3.208627810922665e-05,
"loss": 0.0142,
"step": 285
},
{
"epoch": 1.1002409638554216,
"grad_norm": 0.05139852315187454,
"learning_rate": 3.201484528475958e-05,
"loss": 0.0053,
"step": 286
},
{
"epoch": 1.1040963855421686,
"grad_norm": 0.10800574719905853,
"learning_rate": 3.194317183405573e-05,
"loss": 0.0082,
"step": 287
},
{
"epoch": 1.1079518072289156,
"grad_norm": 0.08289000391960144,
"learning_rate": 3.187125919254869e-05,
"loss": 0.0058,
"step": 288
},
{
"epoch": 1.1118072289156626,
"grad_norm": 0.2032063752412796,
"learning_rate": 3.1799108800462466e-05,
"loss": 0.0081,
"step": 289
},
{
"epoch": 1.1156626506024097,
"grad_norm": 0.09905604273080826,
"learning_rate": 3.172672210278254e-05,
"loss": 0.0094,
"step": 290
},
{
"epoch": 1.1195180722891567,
"grad_norm": 0.13566920161247253,
"learning_rate": 3.1654100549227024e-05,
"loss": 0.0065,
"step": 291
},
{
"epoch": 1.1233734939759037,
"grad_norm": 0.13147567212581635,
"learning_rate": 3.158124559421758e-05,
"loss": 0.0056,
"step": 292
},
{
"epoch": 1.1272289156626507,
"grad_norm": 0.0829431414604187,
"learning_rate": 3.1508158696850275e-05,
"loss": 0.0066,
"step": 293
},
{
"epoch": 1.1310843373493975,
"grad_norm": 0.152206689119339,
"learning_rate": 3.14348413208664e-05,
"loss": 0.0072,
"step": 294
},
{
"epoch": 1.1349397590361445,
"grad_norm": 0.10874905437231064,
"learning_rate": 3.136129493462312e-05,
"loss": 0.0071,
"step": 295
},
{
"epoch": 1.1387951807228915,
"grad_norm": 0.08331774920225143,
"learning_rate": 3.12875210110641e-05,
"loss": 0.0079,
"step": 296
},
{
"epoch": 1.1426506024096386,
"grad_norm": 0.08060551434755325,
"learning_rate": 3.121352102768998e-05,
"loss": 0.0059,
"step": 297
},
{
"epoch": 1.1465060240963856,
"grad_norm": 0.08402010798454285,
"learning_rate": 3.113929646652879e-05,
"loss": 0.0047,
"step": 298
},
{
"epoch": 1.1503614457831326,
"grad_norm": 0.07346951961517334,
"learning_rate": 3.106484881410628e-05,
"loss": 0.0071,
"step": 299
},
{
"epoch": 1.1542168674698796,
"grad_norm": 0.11735919117927551,
"learning_rate": 3.0990179561416124e-05,
"loss": 0.0077,
"step": 300
},
{
"epoch": 1.1580722891566264,
"grad_norm": 0.05967315658926964,
"learning_rate": 3.091529020389009e-05,
"loss": 0.0036,
"step": 301
},
{
"epoch": 1.1619277108433734,
"grad_norm": 2.5192465782165527,
"learning_rate": 3.0840182241368083e-05,
"loss": 0.0089,
"step": 302
},
{
"epoch": 1.1657831325301204,
"grad_norm": 0.12357142567634583,
"learning_rate": 3.076485717806808e-05,
"loss": 0.011,
"step": 303
},
{
"epoch": 1.1696385542168675,
"grad_norm": 0.08216772973537445,
"learning_rate": 3.0689316522556026e-05,
"loss": 0.0147,
"step": 304
},
{
"epoch": 1.1734939759036145,
"grad_norm": 0.10057099908590317,
"learning_rate": 3.061356178771564e-05,
"loss": 0.0143,
"step": 305
},
{
"epoch": 1.1773493975903615,
"grad_norm": 0.14141112565994263,
"learning_rate": 3.053759449071807e-05,
"loss": 0.0104,
"step": 306
},
{
"epoch": 1.1812048192771085,
"grad_norm": 0.08545457571744919,
"learning_rate": 3.0461416152991555e-05,
"loss": 0.0076,
"step": 307
},
{
"epoch": 1.1850602409638555,
"grad_norm": 0.11981064826250076,
"learning_rate": 3.038502830019092e-05,
"loss": 0.0078,
"step": 308
},
{
"epoch": 1.1889156626506023,
"grad_norm": 0.07742572575807571,
"learning_rate": 3.0308432462167045e-05,
"loss": 0.0054,
"step": 309
},
{
"epoch": 1.1927710843373494,
"grad_norm": 0.08275239914655685,
"learning_rate": 3.0231630172936205e-05,
"loss": 0.0053,
"step": 310
},
{
"epoch": 1.1966265060240964,
"grad_norm": 0.42046189308166504,
"learning_rate": 3.015462297064936e-05,
"loss": 0.0258,
"step": 311
},
{
"epoch": 1.2004819277108434,
"grad_norm": 0.07977071404457092,
"learning_rate": 3.0077412397561345e-05,
"loss": 0.0153,
"step": 312
},
{
"epoch": 1.2043373493975904,
"grad_norm": 0.09960279613733292,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.0101,
"step": 313
},
{
"epoch": 1.2081927710843374,
"grad_norm": 0.09214231371879578,
"learning_rate": 2.992238732833518e-05,
"loss": 0.0053,
"step": 314
},
{
"epoch": 1.2120481927710842,
"grad_norm": 0.06532324850559235,
"learning_rate": 2.98445759369477e-05,
"loss": 0.0082,
"step": 315
},
{
"epoch": 1.2159036144578312,
"grad_norm": 0.06730500608682632,
"learning_rate": 2.9766567384198228e-05,
"loss": 0.0063,
"step": 316
},
{
"epoch": 1.2197590361445783,
"grad_norm": 0.1506088674068451,
"learning_rate": 2.9688363232396056e-05,
"loss": 0.0095,
"step": 317
},
{
"epoch": 1.2236144578313253,
"grad_norm": 0.07647290825843811,
"learning_rate": 2.960996504776783e-05,
"loss": 0.0082,
"step": 318
},
{
"epoch": 1.2274698795180723,
"grad_norm": 0.21594130992889404,
"learning_rate": 2.9531374400426158e-05,
"loss": 0.0251,
"step": 319
},
{
"epoch": 1.2313253012048193,
"grad_norm": 0.1463146209716797,
"learning_rate": 2.945259286433819e-05,
"loss": 0.0087,
"step": 320
},
{
"epoch": 1.2351807228915663,
"grad_norm": 0.08222866803407669,
"learning_rate": 2.9373622017294075e-05,
"loss": 0.0062,
"step": 321
},
{
"epoch": 1.2390361445783133,
"grad_norm": 0.09863219410181046,
"learning_rate": 2.9294463440875375e-05,
"loss": 0.0067,
"step": 322
},
{
"epoch": 1.2428915662650604,
"grad_norm": 0.13434955477714539,
"learning_rate": 2.9215118720423375e-05,
"loss": 0.0073,
"step": 323
},
{
"epoch": 1.2467469879518072,
"grad_norm": 0.08647848665714264,
"learning_rate": 2.913558944500736e-05,
"loss": 0.006,
"step": 324
},
{
"epoch": 1.2506024096385542,
"grad_norm": 0.07066208124160767,
"learning_rate": 2.9055877207392752e-05,
"loss": 0.0041,
"step": 325
},
{
"epoch": 1.2544578313253012,
"grad_norm": 0.09136916697025299,
"learning_rate": 2.897598360400925e-05,
"loss": 0.0055,
"step": 326
},
{
"epoch": 1.2583132530120482,
"grad_norm": 0.12554176151752472,
"learning_rate": 2.8895910234918828e-05,
"loss": 0.0117,
"step": 327
},
{
"epoch": 1.2621686746987952,
"grad_norm": 0.1485210657119751,
"learning_rate": 2.8815658703783715e-05,
"loss": 0.0039,
"step": 328
},
{
"epoch": 1.266024096385542,
"grad_norm": 0.09695762395858765,
"learning_rate": 2.873523061783426e-05,
"loss": 0.0091,
"step": 329
},
{
"epoch": 1.269879518072289,
"grad_norm": 0.07787121087312698,
"learning_rate": 2.8654627587836724e-05,
"loss": 0.0041,
"step": 330
},
{
"epoch": 1.273734939759036,
"grad_norm": 0.06576529145240784,
"learning_rate": 2.8573851228061084e-05,
"loss": 0.0041,
"step": 331
},
{
"epoch": 1.277590361445783,
"grad_norm": 0.06353545188903809,
"learning_rate": 2.8492903156248642e-05,
"loss": 0.0071,
"step": 332
},
{
"epoch": 1.28144578313253,
"grad_norm": 0.1047273725271225,
"learning_rate": 2.8411784993579633e-05,
"loss": 0.009,
"step": 333
},
{
"epoch": 1.2853012048192771,
"grad_norm": 0.1017284020781517,
"learning_rate": 2.8330498364640803e-05,
"loss": 0.01,
"step": 334
},
{
"epoch": 1.2891566265060241,
"grad_norm": 0.13284249603748322,
"learning_rate": 2.8249044897392814e-05,
"loss": 0.0124,
"step": 335
},
{
"epoch": 1.2930120481927712,
"grad_norm": 0.08927687257528305,
"learning_rate": 2.8167426223137682e-05,
"loss": 0.0193,
"step": 336
},
{
"epoch": 1.2968674698795182,
"grad_norm": 0.0922897607088089,
"learning_rate": 2.80856439764861e-05,
"loss": 0.0403,
"step": 337
},
{
"epoch": 1.3007228915662652,
"grad_norm": 0.3542531430721283,
"learning_rate": 2.8003699795324674e-05,
"loss": 0.0053,
"step": 338
},
{
"epoch": 1.304578313253012,
"grad_norm": 0.10456252843141556,
"learning_rate": 2.792159532078314e-05,
"loss": 0.0126,
"step": 339
},
{
"epoch": 1.308433734939759,
"grad_norm": 0.06387709081172943,
"learning_rate": 2.7839332197201505e-05,
"loss": 0.0036,
"step": 340
},
{
"epoch": 1.312289156626506,
"grad_norm": 0.06973070651292801,
"learning_rate": 2.77569120720971e-05,
"loss": 0.0094,
"step": 341
},
{
"epoch": 1.316144578313253,
"grad_norm": 0.08025919646024704,
"learning_rate": 2.7674336596131596e-05,
"loss": 0.0081,
"step": 342
},
{
"epoch": 1.32,
"grad_norm": 0.15938566625118256,
"learning_rate": 2.7591607423077932e-05,
"loss": 0.0104,
"step": 343
},
{
"epoch": 1.3238554216867469,
"grad_norm": 0.13838085532188416,
"learning_rate": 2.75087262097872e-05,
"loss": 0.004,
"step": 344
},
{
"epoch": 1.3277108433734939,
"grad_norm": 0.06210877001285553,
"learning_rate": 2.7425694616155474e-05,
"loss": 0.0044,
"step": 345
},
{
"epoch": 1.331566265060241,
"grad_norm": 0.094986192882061,
"learning_rate": 2.7342514305090554e-05,
"loss": 0.0206,
"step": 346
},
{
"epoch": 1.335421686746988,
"grad_norm": 0.07161661237478256,
"learning_rate": 2.7259186942478656e-05,
"loss": 0.0033,
"step": 347
},
{
"epoch": 1.339277108433735,
"grad_norm": 0.12837430834770203,
"learning_rate": 2.717571419715107e-05,
"loss": 0.0143,
"step": 348
},
{
"epoch": 1.343132530120482,
"grad_norm": 0.07206165045499802,
"learning_rate": 2.7092097740850712e-05,
"loss": 0.0037,
"step": 349
},
{
"epoch": 1.346987951807229,
"grad_norm": 0.06978686898946762,
"learning_rate": 2.700833924819868e-05,
"loss": 0.0059,
"step": 350
},
{
"epoch": 1.350843373493976,
"grad_norm": 0.14167740941047668,
"learning_rate": 2.692444039666066e-05,
"loss": 0.0093,
"step": 351
},
{
"epoch": 1.354698795180723,
"grad_norm": 0.08207441866397858,
"learning_rate": 2.684040286651338e-05,
"loss": 0.0087,
"step": 352
},
{
"epoch": 1.3585542168674698,
"grad_norm": 0.08811763674020767,
"learning_rate": 2.6756228340810946e-05,
"loss": 0.0076,
"step": 353
},
{
"epoch": 1.3624096385542168,
"grad_norm": 0.09116017818450928,
"learning_rate": 2.6671918505351132e-05,
"loss": 0.0043,
"step": 354
},
{
"epoch": 1.3662650602409638,
"grad_norm": 0.06122549995779991,
"learning_rate": 2.6587475048641596e-05,
"loss": 0.0032,
"step": 355
},
{
"epoch": 1.3701204819277109,
"grad_norm": 0.07415001839399338,
"learning_rate": 2.6502899661866104e-05,
"loss": 0.0072,
"step": 356
},
{
"epoch": 1.3739759036144579,
"grad_norm": 0.08215410262346268,
"learning_rate": 2.6418194038850634e-05,
"loss": 0.009,
"step": 357
},
{
"epoch": 1.377831325301205,
"grad_norm": 0.0975186750292778,
"learning_rate": 2.6333359876029455e-05,
"loss": 0.0045,
"step": 358
},
{
"epoch": 1.3816867469879517,
"grad_norm": 0.20146572589874268,
"learning_rate": 2.624839887241115e-05,
"loss": 0.0094,
"step": 359
},
{
"epoch": 1.3855421686746987,
"grad_norm": 0.08265519142150879,
"learning_rate": 2.6163312729544616e-05,
"loss": 0.0047,
"step": 360
},
{
"epoch": 1.3893975903614457,
"grad_norm": 0.09653282910585403,
"learning_rate": 2.607810315148494e-05,
"loss": 0.0077,
"step": 361
},
{
"epoch": 1.3932530120481927,
"grad_norm": 0.10554955899715424,
"learning_rate": 2.5992771844759315e-05,
"loss": 0.0091,
"step": 362
},
{
"epoch": 1.3971084337349398,
"grad_norm": 0.06274604052305222,
"learning_rate": 2.5907320518332827e-05,
"loss": 0.0073,
"step": 363
},
{
"epoch": 1.4009638554216868,
"grad_norm": 0.07398885488510132,
"learning_rate": 2.582175088357426e-05,
"loss": 0.0122,
"step": 364
},
{
"epoch": 1.4048192771084338,
"grad_norm": 0.0733281746506691,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.0107,
"step": 365
},
{
"epoch": 1.4086746987951808,
"grad_norm": 0.08075340837240219,
"learning_rate": 2.5650263546348743e-05,
"loss": 0.0071,
"step": 366
},
{
"epoch": 1.4125301204819278,
"grad_norm": 0.22409602999687195,
"learning_rate": 2.5564349278329056e-05,
"loss": 0.0078,
"step": 367
},
{
"epoch": 1.4163855421686746,
"grad_norm": 0.07538998126983643,
"learning_rate": 2.547832357080305e-05,
"loss": 0.0115,
"step": 368
},
{
"epoch": 1.4202409638554216,
"grad_norm": 0.13440506160259247,
"learning_rate": 2.539218814664288e-05,
"loss": 0.0097,
"step": 369
},
{
"epoch": 1.4240963855421687,
"grad_norm": 0.17739734053611755,
"learning_rate": 2.5305944730918017e-05,
"loss": 0.0151,
"step": 370
},
{
"epoch": 1.4279518072289157,
"grad_norm": 0.07818708568811417,
"learning_rate": 2.521959505086075e-05,
"loss": 0.0048,
"step": 371
},
{
"epoch": 1.4318072289156627,
"grad_norm": 0.1267346739768982,
"learning_rate": 2.5133140835831537e-05,
"loss": 0.0123,
"step": 372
},
{
"epoch": 1.4356626506024097,
"grad_norm": 0.06766249984502792,
"learning_rate": 2.5046583817284437e-05,
"loss": 0.0092,
"step": 373
},
{
"epoch": 1.4395180722891565,
"grad_norm": 0.08244456350803375,
"learning_rate": 2.495992572873237e-05,
"loss": 0.0055,
"step": 374
},
{
"epoch": 1.4433734939759035,
"grad_norm": 0.15402929484844208,
"learning_rate": 2.487316830571244e-05,
"loss": 0.0115,
"step": 375
},
{
"epoch": 1.4472289156626506,
"grad_norm": 0.055807530879974365,
"learning_rate": 2.4786313285751158e-05,
"loss": 0.005,
"step": 376
},
{
"epoch": 1.4510843373493976,
"grad_norm": 0.26038140058517456,
"learning_rate": 2.4699362408329646e-05,
"loss": 0.0047,
"step": 377
},
{
"epoch": 1.4549397590361446,
"grad_norm": 0.0664365142583847,
"learning_rate": 2.4612317414848804e-05,
"loss": 0.0047,
"step": 378
},
{
"epoch": 1.4587951807228916,
"grad_norm": 5.076003551483154,
"learning_rate": 2.4525180048594452e-05,
"loss": 0.004,
"step": 379
},
{
"epoch": 1.4626506024096386,
"grad_norm": 0.0970849022269249,
"learning_rate": 2.4437952054702354e-05,
"loss": 0.0074,
"step": 380
},
{
"epoch": 1.4665060240963856,
"grad_norm": 0.19724144041538239,
"learning_rate": 2.435063518012335e-05,
"loss": 0.014,
"step": 381
},
{
"epoch": 1.4703614457831327,
"grad_norm": 0.07128898799419403,
"learning_rate": 2.4263231173588323e-05,
"loss": 0.0071,
"step": 382
},
{
"epoch": 1.4742168674698795,
"grad_norm": 0.4438879191875458,
"learning_rate": 2.4175741785573177e-05,
"loss": 0.0096,
"step": 383
},
{
"epoch": 1.4780722891566265,
"grad_norm": 0.0698755756020546,
"learning_rate": 2.4088168768263782e-05,
"loss": 0.0068,
"step": 384
},
{
"epoch": 1.4819277108433735,
"grad_norm": 0.10384954512119293,
"learning_rate": 2.4000513875520892e-05,
"loss": 0.0236,
"step": 385
},
{
"epoch": 1.4857831325301205,
"grad_norm": 0.0846671611070633,
"learning_rate": 2.3912778862845015e-05,
"loss": 0.0076,
"step": 386
},
{
"epoch": 1.4896385542168675,
"grad_norm": 0.08970258384943008,
"learning_rate": 2.3824965487341247e-05,
"loss": 0.0069,
"step": 387
},
{
"epoch": 1.4934939759036143,
"grad_norm": 0.0815376341342926,
"learning_rate": 2.3737075507684103e-05,
"loss": 0.0083,
"step": 388
},
{
"epoch": 1.4973493975903613,
"grad_norm": 0.05652186647057533,
"learning_rate": 2.3649110684082258e-05,
"loss": 0.0034,
"step": 389
},
{
"epoch": 1.5012048192771084,
"grad_norm": 0.09172943234443665,
"learning_rate": 2.3561072778243335e-05,
"loss": 0.0095,
"step": 390
},
{
"epoch": 1.5050602409638554,
"grad_norm": 0.07991674542427063,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.0054,
"step": 391
},
{
"epoch": 1.5089156626506024,
"grad_norm": 0.06812671571969986,
"learning_rate": 2.3384784773967675e-05,
"loss": 0.0127,
"step": 392
},
{
"epoch": 1.5127710843373494,
"grad_norm": 0.07687579840421677,
"learning_rate": 2.3296538206123134e-05,
"loss": 0.004,
"step": 393
},
{
"epoch": 1.5166265060240964,
"grad_norm": 0.09379265457391739,
"learning_rate": 2.3208225617155206e-05,
"loss": 0.0086,
"step": 394
},
{
"epoch": 1.5204819277108435,
"grad_norm": 0.16897223889827728,
"learning_rate": 2.311984877573636e-05,
"loss": 0.013,
"step": 395
},
{
"epoch": 1.5243373493975905,
"grad_norm": 0.9250982403755188,
"learning_rate": 2.3031409451825852e-05,
"loss": 0.0181,
"step": 396
},
{
"epoch": 1.5281927710843375,
"grad_norm": 0.21858066320419312,
"learning_rate": 2.2942909416634326e-05,
"loss": 0.0051,
"step": 397
},
{
"epoch": 1.5320481927710843,
"grad_norm": 0.07181593775749207,
"learning_rate": 2.285435044258829e-05,
"loss": 0.0062,
"step": 398
},
{
"epoch": 1.5359036144578313,
"grad_norm": 0.16755104064941406,
"learning_rate": 2.2765734303294666e-05,
"loss": 0.0102,
"step": 399
},
{
"epoch": 1.5397590361445783,
"grad_norm": 0.06312964856624603,
"learning_rate": 2.2677062773505235e-05,
"loss": 0.0075,
"step": 400
},
{
"epoch": 1.5436144578313253,
"grad_norm": 0.06663095206022263,
"learning_rate": 2.2588337629081107e-05,
"loss": 0.0077,
"step": 401
},
{
"epoch": 1.5474698795180721,
"grad_norm": 0.05746052786707878,
"learning_rate": 2.249956064695717e-05,
"loss": 0.0053,
"step": 402
},
{
"epoch": 1.5513253012048192,
"grad_norm": 0.0769314169883728,
"learning_rate": 2.2410733605106462e-05,
"loss": 0.0047,
"step": 403
},
{
"epoch": 1.5551807228915662,
"grad_norm": 0.1205008402466774,
"learning_rate": 2.2321858282504606e-05,
"loss": 0.0159,
"step": 404
},
{
"epoch": 1.5590361445783132,
"grad_norm": 0.08342281728982925,
"learning_rate": 2.2232936459094158e-05,
"loss": 0.0082,
"step": 405
},
{
"epoch": 1.5628915662650602,
"grad_norm": 0.358065128326416,
"learning_rate": 2.214396991574896e-05,
"loss": 0.0126,
"step": 406
},
{
"epoch": 1.5667469879518072,
"grad_norm": 0.05992039665579796,
"learning_rate": 2.205496043423849e-05,
"loss": 0.0039,
"step": 407
},
{
"epoch": 1.5706024096385542,
"grad_norm": 0.07799182087182999,
"learning_rate": 2.1965909797192143e-05,
"loss": 0.005,
"step": 408
},
{
"epoch": 1.5744578313253013,
"grad_norm": 0.09712725132703781,
"learning_rate": 2.1876819788063586e-05,
"loss": 0.006,
"step": 409
},
{
"epoch": 1.5783132530120483,
"grad_norm": 0.05882284417748451,
"learning_rate": 2.1787692191094986e-05,
"loss": 0.0053,
"step": 410
},
{
"epoch": 1.5821686746987953,
"grad_norm": 0.06840016692876816,
"learning_rate": 2.16985287912813e-05,
"loss": 0.006,
"step": 411
},
{
"epoch": 1.5860240963855423,
"grad_norm": 0.07110153138637543,
"learning_rate": 2.1609331374334523e-05,
"loss": 0.0049,
"step": 412
},
{
"epoch": 1.589879518072289,
"grad_norm": 0.1433626413345337,
"learning_rate": 2.1520101726647922e-05,
"loss": 0.0115,
"step": 413
},
{
"epoch": 1.5937349397590361,
"grad_norm": 0.07788779586553574,
"learning_rate": 2.1430841635260274e-05,
"loss": 0.0071,
"step": 414
},
{
"epoch": 1.5975903614457831,
"grad_norm": 0.06472036987543106,
"learning_rate": 2.1341552887820048e-05,
"loss": 0.0095,
"step": 415
},
{
"epoch": 1.6014457831325302,
"grad_norm": 0.10168538242578506,
"learning_rate": 2.125223727254964e-05,
"loss": 0.0162,
"step": 416
},
{
"epoch": 1.605301204819277,
"grad_norm": 0.098752461373806,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.0147,
"step": 417
},
{
"epoch": 1.609156626506024,
"grad_norm": 0.0626005232334137,
"learning_rate": 2.1073532594062432e-05,
"loss": 0.0053,
"step": 418
},
{
"epoch": 1.613012048192771,
"grad_norm": 0.07270153611898422,
"learning_rate": 2.0984147109837564e-05,
"loss": 0.0055,
"step": 419
},
{
"epoch": 1.616867469879518,
"grad_norm": 0.06415501236915588,
"learning_rate": 2.0894741915694685e-05,
"loss": 0.0036,
"step": 420
},
{
"epoch": 1.620722891566265,
"grad_norm": 0.11951615661382675,
"learning_rate": 2.0805318802188307e-05,
"loss": 0.0082,
"step": 421
},
{
"epoch": 1.624578313253012,
"grad_norm": 0.0676022469997406,
"learning_rate": 2.071587956023181e-05,
"loss": 0.0036,
"step": 422
},
{
"epoch": 1.628433734939759,
"grad_norm": 0.05662338435649872,
"learning_rate": 2.0626425981061608e-05,
"loss": 0.0048,
"step": 423
},
{
"epoch": 1.632289156626506,
"grad_norm": 0.07476495206356049,
"learning_rate": 2.053695985620123e-05,
"loss": 0.0045,
"step": 424
},
{
"epoch": 1.636144578313253,
"grad_norm": 0.07783601433038712,
"learning_rate": 2.0447482977425465e-05,
"loss": 0.0059,
"step": 425
},
{
"epoch": 1.6400000000000001,
"grad_norm": 0.06224929541349411,
"learning_rate": 2.0357997136724486e-05,
"loss": 0.0151,
"step": 426
},
{
"epoch": 1.6438554216867471,
"grad_norm": 0.06190069392323494,
"learning_rate": 2.0268504126267952e-05,
"loss": 0.0047,
"step": 427
},
{
"epoch": 1.647710843373494,
"grad_norm": 0.05491878464818001,
"learning_rate": 2.0179005738369098e-05,
"loss": 0.0066,
"step": 428
},
{
"epoch": 1.651566265060241,
"grad_norm": 0.10569415986537933,
"learning_rate": 2.008950376544887e-05,
"loss": 0.0068,
"step": 429
},
{
"epoch": 1.655421686746988,
"grad_norm": 0.08094307035207748,
"learning_rate": 2e-05,
"loss": 0.0084,
"step": 430
},
{
"epoch": 1.659277108433735,
"grad_norm": 0.055527571588754654,
"learning_rate": 1.9910496234551132e-05,
"loss": 0.0053,
"step": 431
},
{
"epoch": 1.6631325301204818,
"grad_norm": 0.0911506935954094,
"learning_rate": 1.982099426163091e-05,
"loss": 0.0067,
"step": 432
},
{
"epoch": 1.6669879518072288,
"grad_norm": 0.09402691572904587,
"learning_rate": 1.9731495873732055e-05,
"loss": 0.0065,
"step": 433
},
{
"epoch": 1.6708433734939758,
"grad_norm": 0.07658103108406067,
"learning_rate": 1.9642002863275517e-05,
"loss": 0.0121,
"step": 434
},
{
"epoch": 1.6746987951807228,
"grad_norm": 0.07842761278152466,
"learning_rate": 1.9552517022574542e-05,
"loss": 0.0097,
"step": 435
},
{
"epoch": 1.6785542168674699,
"grad_norm": 0.14522653818130493,
"learning_rate": 1.946304014379878e-05,
"loss": 0.0068,
"step": 436
},
{
"epoch": 1.6824096385542169,
"grad_norm": 0.17214494943618774,
"learning_rate": 1.93735740189384e-05,
"loss": 0.0161,
"step": 437
},
{
"epoch": 1.686265060240964,
"grad_norm": 0.08584551513195038,
"learning_rate": 1.9284120439768192e-05,
"loss": 0.0118,
"step": 438
},
{
"epoch": 1.690120481927711,
"grad_norm": 0.057091426104307175,
"learning_rate": 1.9194681197811703e-05,
"loss": 0.0096,
"step": 439
},
{
"epoch": 1.693975903614458,
"grad_norm": 0.1121736541390419,
"learning_rate": 1.910525808430532e-05,
"loss": 0.0066,
"step": 440
},
{
"epoch": 1.697831325301205,
"grad_norm": 0.05429166927933693,
"learning_rate": 1.901585289016244e-05,
"loss": 0.0033,
"step": 441
},
{
"epoch": 1.701686746987952,
"grad_norm": 0.06648131459951401,
"learning_rate": 1.8926467405937567e-05,
"loss": 0.0074,
"step": 442
},
{
"epoch": 1.7055421686746988,
"grad_norm": 0.124962218105793,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.02,
"step": 443
},
{
"epoch": 1.7093975903614458,
"grad_norm": 0.08059555292129517,
"learning_rate": 1.8747762727450363e-05,
"loss": 0.0126,
"step": 444
},
{
"epoch": 1.7132530120481928,
"grad_norm": 0.11686407029628754,
"learning_rate": 1.8658447112179952e-05,
"loss": 0.0245,
"step": 445
},
{
"epoch": 1.7171084337349396,
"grad_norm": 0.047245193272829056,
"learning_rate": 1.8569158364739736e-05,
"loss": 0.0028,
"step": 446
},
{
"epoch": 1.7209638554216866,
"grad_norm": 0.06443202495574951,
"learning_rate": 1.8479898273352084e-05,
"loss": 0.004,
"step": 447
},
{
"epoch": 1.7248192771084336,
"grad_norm": 0.045282021164894104,
"learning_rate": 1.8390668625665483e-05,
"loss": 0.003,
"step": 448
},
{
"epoch": 1.7286746987951807,
"grad_norm": 0.13053859770298004,
"learning_rate": 1.83014712087187e-05,
"loss": 0.0102,
"step": 449
},
{
"epoch": 1.7325301204819277,
"grad_norm": 0.07346352934837341,
"learning_rate": 1.821230780890502e-05,
"loss": 0.0089,
"step": 450
},
{
"epoch": 1.7363855421686747,
"grad_norm": 0.08662671595811844,
"learning_rate": 1.8123180211936417e-05,
"loss": 0.0127,
"step": 451
},
{
"epoch": 1.7402409638554217,
"grad_norm": 0.04067783057689667,
"learning_rate": 1.803409020280786e-05,
"loss": 0.0042,
"step": 452
},
{
"epoch": 1.7440963855421687,
"grad_norm": 0.04167520999908447,
"learning_rate": 1.794503956576152e-05,
"loss": 0.0028,
"step": 453
},
{
"epoch": 1.7479518072289157,
"grad_norm": 0.0644107311964035,
"learning_rate": 1.7856030084251045e-05,
"loss": 0.0089,
"step": 454
},
{
"epoch": 1.7518072289156628,
"grad_norm": 0.1271977722644806,
"learning_rate": 1.776706354090585e-05,
"loss": 0.0081,
"step": 455
},
{
"epoch": 1.7556626506024098,
"grad_norm": 0.04880126565694809,
"learning_rate": 1.7678141717495394e-05,
"loss": 0.0037,
"step": 456
},
{
"epoch": 1.7595180722891566,
"grad_norm": 0.07817039638757706,
"learning_rate": 1.758926639489354e-05,
"loss": 0.0156,
"step": 457
},
{
"epoch": 1.7633734939759036,
"grad_norm": 0.0569678358733654,
"learning_rate": 1.7500439353042834e-05,
"loss": 0.011,
"step": 458
},
{
"epoch": 1.7672289156626506,
"grad_norm": 0.08766577392816544,
"learning_rate": 1.7411662370918893e-05,
"loss": 0.004,
"step": 459
},
{
"epoch": 1.7710843373493976,
"grad_norm": 0.07262749969959259,
"learning_rate": 1.7322937226494775e-05,
"loss": 0.0053,
"step": 460
},
{
"epoch": 1.7749397590361444,
"grad_norm": 0.07411035895347595,
"learning_rate": 1.7234265696705344e-05,
"loss": 0.0047,
"step": 461
},
{
"epoch": 1.7787951807228914,
"grad_norm": 0.047550980001688004,
"learning_rate": 1.7145649557411716e-05,
"loss": 0.0036,
"step": 462
},
{
"epoch": 1.7826506024096385,
"grad_norm": 0.06966912746429443,
"learning_rate": 1.7057090583365678e-05,
"loss": 0.0107,
"step": 463
},
{
"epoch": 1.7865060240963855,
"grad_norm": 0.09174683690071106,
"learning_rate": 1.6968590548174155e-05,
"loss": 0.0047,
"step": 464
},
{
"epoch": 1.7903614457831325,
"grad_norm": 0.10760336369276047,
"learning_rate": 1.6880151224263646e-05,
"loss": 0.0107,
"step": 465
},
{
"epoch": 1.7942168674698795,
"grad_norm": 0.08315455913543701,
"learning_rate": 1.6791774382844794e-05,
"loss": 0.0097,
"step": 466
},
{
"epoch": 1.7980722891566265,
"grad_norm": 0.044719114899635315,
"learning_rate": 1.6703461793876876e-05,
"loss": 0.0052,
"step": 467
},
{
"epoch": 1.8019277108433736,
"grad_norm": 0.06534062325954437,
"learning_rate": 1.6615215226032332e-05,
"loss": 0.0063,
"step": 468
},
{
"epoch": 1.8057831325301206,
"grad_norm": 0.07542230188846588,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.0112,
"step": 469
},
{
"epoch": 1.8096385542168676,
"grad_norm": 0.07965763658285141,
"learning_rate": 1.643892722175667e-05,
"loss": 0.0103,
"step": 470
},
{
"epoch": 1.8134939759036146,
"grad_norm": 0.0694553405046463,
"learning_rate": 1.635088931591775e-05,
"loss": 0.0054,
"step": 471
},
{
"epoch": 1.8173493975903614,
"grad_norm": 0.16062794625759125,
"learning_rate": 1.6262924492315907e-05,
"loss": 0.0273,
"step": 472
},
{
"epoch": 1.8212048192771084,
"grad_norm": 0.06760603934526443,
"learning_rate": 1.6175034512658753e-05,
"loss": 0.0051,
"step": 473
},
{
"epoch": 1.8250602409638554,
"grad_norm": 0.07885128259658813,
"learning_rate": 1.6087221137154995e-05,
"loss": 0.014,
"step": 474
},
{
"epoch": 1.8289156626506025,
"grad_norm": 0.10484739392995834,
"learning_rate": 1.5999486124479115e-05,
"loss": 0.0146,
"step": 475
},
{
"epoch": 1.8327710843373493,
"grad_norm": 0.05339181050658226,
"learning_rate": 1.591183123173622e-05,
"loss": 0.0055,
"step": 476
},
{
"epoch": 1.8366265060240963,
"grad_norm": 0.05444730818271637,
"learning_rate": 1.5824258214426833e-05,
"loss": 0.0045,
"step": 477
},
{
"epoch": 1.8404819277108433,
"grad_norm": 0.07196304202079773,
"learning_rate": 1.5736768826411683e-05,
"loss": 0.0062,
"step": 478
},
{
"epoch": 1.8443373493975903,
"grad_norm": 0.08023171871900558,
"learning_rate": 1.5649364819876655e-05,
"loss": 0.01,
"step": 479
},
{
"epoch": 1.8481927710843373,
"grad_norm": 0.08942972123622894,
"learning_rate": 1.5562047945297652e-05,
"loss": 0.0147,
"step": 480
},
{
"epoch": 1.8520481927710843,
"grad_norm": 0.05932997912168503,
"learning_rate": 1.547481995140556e-05,
"loss": 0.004,
"step": 481
},
{
"epoch": 1.8559036144578314,
"grad_norm": 0.09472596645355225,
"learning_rate": 1.53876825851512e-05,
"loss": 0.0035,
"step": 482
},
{
"epoch": 1.8597590361445784,
"grad_norm": 0.061298541724681854,
"learning_rate": 1.5300637591670357e-05,
"loss": 0.0083,
"step": 483
},
{
"epoch": 1.8636144578313254,
"grad_norm": 0.13098923861980438,
"learning_rate": 1.5213686714248852e-05,
"loss": 0.0095,
"step": 484
},
{
"epoch": 1.8674698795180724,
"grad_norm": 0.08183276653289795,
"learning_rate": 1.5126831694287564e-05,
"loss": 0.0153,
"step": 485
},
{
"epoch": 1.8713253012048194,
"grad_norm": 0.4021635055541992,
"learning_rate": 1.5040074271267633e-05,
"loss": 0.0221,
"step": 486
},
{
"epoch": 1.8751807228915662,
"grad_norm": 0.07240505516529083,
"learning_rate": 1.4953416182715566e-05,
"loss": 0.0035,
"step": 487
},
{
"epoch": 1.8790361445783132,
"grad_norm": 0.15526820719242096,
"learning_rate": 1.4866859164168466e-05,
"loss": 0.0081,
"step": 488
},
{
"epoch": 1.8828915662650603,
"grad_norm": 0.07194899767637253,
"learning_rate": 1.478040494913926e-05,
"loss": 0.0063,
"step": 489
},
{
"epoch": 1.886746987951807,
"grad_norm": 0.07030153274536133,
"learning_rate": 1.4694055269081987e-05,
"loss": 0.0058,
"step": 490
},
{
"epoch": 1.890602409638554,
"grad_norm": 0.1665412038564682,
"learning_rate": 1.460781185335713e-05,
"loss": 0.0061,
"step": 491
},
{
"epoch": 1.894457831325301,
"grad_norm": 0.06069282069802284,
"learning_rate": 1.452167642919695e-05,
"loss": 0.0034,
"step": 492
},
{
"epoch": 1.8983132530120481,
"grad_norm": 0.08425285667181015,
"learning_rate": 1.443565072167095e-05,
"loss": 0.007,
"step": 493
},
{
"epoch": 1.9021686746987951,
"grad_norm": 0.049832381308078766,
"learning_rate": 1.4349736453651262e-05,
"loss": 0.0038,
"step": 494
},
{
"epoch": 1.9060240963855422,
"grad_norm": 0.06115152686834335,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.0057,
"step": 495
},
{
"epoch": 1.9098795180722892,
"grad_norm": 0.07054963707923889,
"learning_rate": 1.4178249116425744e-05,
"loss": 0.0199,
"step": 496
},
{
"epoch": 1.9137349397590362,
"grad_norm": 0.0560605488717556,
"learning_rate": 1.409267948166718e-05,
"loss": 0.0061,
"step": 497
},
{
"epoch": 1.9175903614457832,
"grad_norm": 0.2144383043050766,
"learning_rate": 1.4007228155240696e-05,
"loss": 0.0136,
"step": 498
},
{
"epoch": 1.9214457831325302,
"grad_norm": 0.06980622559785843,
"learning_rate": 1.3921896848515064e-05,
"loss": 0.0096,
"step": 499
},
{
"epoch": 1.9253012048192772,
"grad_norm": 0.07482379674911499,
"learning_rate": 1.3836687270455388e-05,
"loss": 0.0142,
"step": 500
},
{
"epoch": 1.929156626506024,
"grad_norm": 0.053662531077861786,
"learning_rate": 1.3751601127588849e-05,
"loss": 0.005,
"step": 501
},
{
"epoch": 1.933012048192771,
"grad_norm": 0.0578133761882782,
"learning_rate": 1.3666640123970556e-05,
"loss": 0.0042,
"step": 502
},
{
"epoch": 1.936867469879518,
"grad_norm": 0.06373419612646103,
"learning_rate": 1.3581805961149371e-05,
"loss": 0.0091,
"step": 503
},
{
"epoch": 1.940722891566265,
"grad_norm": 0.09418510645627975,
"learning_rate": 1.3497100338133899e-05,
"loss": 0.0162,
"step": 504
},
{
"epoch": 1.944578313253012,
"grad_norm": 0.06826018542051315,
"learning_rate": 1.341252495135841e-05,
"loss": 0.0086,
"step": 505
},
{
"epoch": 1.948433734939759,
"grad_norm": 0.06797529011964798,
"learning_rate": 1.3328081494648874e-05,
"loss": 0.0101,
"step": 506
},
{
"epoch": 1.952289156626506,
"grad_norm": 0.062073782086372375,
"learning_rate": 1.324377165918906e-05,
"loss": 0.0041,
"step": 507
},
{
"epoch": 1.956144578313253,
"grad_norm": 0.05104968696832657,
"learning_rate": 1.3159597133486628e-05,
"loss": 0.0063,
"step": 508
},
{
"epoch": 1.96,
"grad_norm": 0.04420260339975357,
"learning_rate": 1.3075559603339354e-05,
"loss": 0.0037,
"step": 509
},
{
"epoch": 1.963855421686747,
"grad_norm": 0.04095982387661934,
"learning_rate": 1.299166075180133e-05,
"loss": 0.0027,
"step": 510
},
{
"epoch": 1.967710843373494,
"grad_norm": 0.07660691440105438,
"learning_rate": 1.2907902259149287e-05,
"loss": 0.004,
"step": 511
},
{
"epoch": 1.971566265060241,
"grad_norm": 0.17664897441864014,
"learning_rate": 1.282428580284894e-05,
"loss": 0.0068,
"step": 512
},
{
"epoch": 1.975421686746988,
"grad_norm": 0.046913839876651764,
"learning_rate": 1.274081305752135e-05,
"loss": 0.003,
"step": 513
},
{
"epoch": 1.979277108433735,
"grad_norm": 0.056782662868499756,
"learning_rate": 1.2657485694909451e-05,
"loss": 0.0049,
"step": 514
},
{
"epoch": 1.983132530120482,
"grad_norm": 0.07346270233392715,
"learning_rate": 1.2574305383844528e-05,
"loss": 0.0059,
"step": 515
},
{
"epoch": 1.9869879518072289,
"grad_norm": 0.0482415035367012,
"learning_rate": 1.2491273790212806e-05,
"loss": 0.0036,
"step": 516
},
{
"epoch": 1.9908433734939759,
"grad_norm": 0.08729434013366699,
"learning_rate": 1.2408392576922075e-05,
"loss": 0.0084,
"step": 517
},
{
"epoch": 1.994698795180723,
"grad_norm": 0.09798599034547806,
"learning_rate": 1.2325663403868406e-05,
"loss": 0.0104,
"step": 518
},
{
"epoch": 1.99855421686747,
"grad_norm": 0.04882819578051567,
"learning_rate": 1.2243087927902905e-05,
"loss": 0.0028,
"step": 519
},
{
"epoch": 2.0,
"grad_norm": 0.09687007963657379,
"learning_rate": 1.2160667802798503e-05,
"loss": 0.0104,
"step": 520
},
{
"epoch": 2.003855421686747,
"grad_norm": 0.1312517374753952,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.0056,
"step": 521
},
{
"epoch": 2.007710843373494,
"grad_norm": 0.0774284154176712,
"learning_rate": 1.199630020467533e-05,
"loss": 0.0033,
"step": 522
},
{
"epoch": 2.011566265060241,
"grad_norm": 0.12178341299295425,
"learning_rate": 1.1914356023513904e-05,
"loss": 0.0296,
"step": 523
},
{
"epoch": 2.015421686746988,
"grad_norm": 0.0642552599310875,
"learning_rate": 1.1832573776862319e-05,
"loss": 0.0047,
"step": 524
},
{
"epoch": 2.019277108433735,
"grad_norm": 0.03751795366406441,
"learning_rate": 1.1750955102607193e-05,
"loss": 0.0049,
"step": 525
},
{
"epoch": 2.023132530120482,
"grad_norm": 0.05329310894012451,
"learning_rate": 1.1669501635359212e-05,
"loss": 0.0053,
"step": 526
},
{
"epoch": 2.026987951807229,
"grad_norm": 0.049860551953315735,
"learning_rate": 1.1588215006420374e-05,
"loss": 0.0224,
"step": 527
},
{
"epoch": 2.0308433734939757,
"grad_norm": 0.042429227381944656,
"learning_rate": 1.1507096843751372e-05,
"loss": 0.0044,
"step": 528
},
{
"epoch": 2.0346987951807227,
"grad_norm": 0.06820591539144516,
"learning_rate": 1.1426148771938915e-05,
"loss": 0.0074,
"step": 529
},
{
"epoch": 2.0385542168674697,
"grad_norm": 0.03642263635993004,
"learning_rate": 1.1345372412163276e-05,
"loss": 0.0016,
"step": 530
},
{
"epoch": 2.0424096385542168,
"grad_norm": 0.04521489888429642,
"learning_rate": 1.1264769382165748e-05,
"loss": 0.0028,
"step": 531
},
{
"epoch": 2.0462650602409638,
"grad_norm": 0.052644696086645126,
"learning_rate": 1.1184341296216281e-05,
"loss": 0.0025,
"step": 532
},
{
"epoch": 2.050120481927711,
"grad_norm": 0.04841390997171402,
"learning_rate": 1.110408976508118e-05,
"loss": 0.0022,
"step": 533
},
{
"epoch": 2.053975903614458,
"grad_norm": 0.03452732786536217,
"learning_rate": 1.1024016395990758e-05,
"loss": 0.0025,
"step": 534
},
{
"epoch": 2.057831325301205,
"grad_norm": 0.07203777134418488,
"learning_rate": 1.094412279260726e-05,
"loss": 0.0072,
"step": 535
},
{
"epoch": 2.061686746987952,
"grad_norm": 0.04625348746776581,
"learning_rate": 1.0864410554992654e-05,
"loss": 0.0026,
"step": 536
},
{
"epoch": 2.065542168674699,
"grad_norm": 0.03372524306178093,
"learning_rate": 1.0784881279576635e-05,
"loss": 0.0014,
"step": 537
},
{
"epoch": 2.069397590361446,
"grad_norm": 0.043473415076732635,
"learning_rate": 1.070553655912463e-05,
"loss": 0.0028,
"step": 538
},
{
"epoch": 2.073253012048193,
"grad_norm": 0.07593965530395508,
"learning_rate": 1.0626377982705929e-05,
"loss": 0.0034,
"step": 539
},
{
"epoch": 2.07710843373494,
"grad_norm": 0.23622868955135345,
"learning_rate": 1.0547407135661816e-05,
"loss": 0.0208,
"step": 540
},
{
"epoch": 2.080963855421687,
"grad_norm": 0.06061263382434845,
"learning_rate": 1.0468625599573842e-05,
"loss": 0.0033,
"step": 541
},
{
"epoch": 2.0848192771084335,
"grad_norm": 0.04391901195049286,
"learning_rate": 1.0390034952232178e-05,
"loss": 0.0013,
"step": 542
},
{
"epoch": 2.0886746987951805,
"grad_norm": 0.03637627512216568,
"learning_rate": 1.0311636767603952e-05,
"loss": 0.002,
"step": 543
},
{
"epoch": 2.0925301204819275,
"grad_norm": 0.1552000641822815,
"learning_rate": 1.0233432615801786e-05,
"loss": 0.0134,
"step": 544
},
{
"epoch": 2.0963855421686746,
"grad_norm": 0.03941213712096214,
"learning_rate": 1.0155424063052306e-05,
"loss": 0.0021,
"step": 545
},
{
"epoch": 2.1002409638554216,
"grad_norm": 0.06914766877889633,
"learning_rate": 1.0077612671664827e-05,
"loss": 0.0063,
"step": 546
},
{
"epoch": 2.1040963855421686,
"grad_norm": 0.07488182187080383,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.011,
"step": 547
},
{
"epoch": 2.1079518072289156,
"grad_norm": 0.0428319126367569,
"learning_rate": 9.922587602438657e-06,
"loss": 0.0048,
"step": 548
},
{
"epoch": 2.1118072289156626,
"grad_norm": 0.03909826651215553,
"learning_rate": 9.84537702935065e-06,
"loss": 0.0019,
"step": 549
},
{
"epoch": 2.1156626506024097,
"grad_norm": 0.05299150571227074,
"learning_rate": 9.768369827063803e-06,
"loss": 0.0075,
"step": 550
},
{
"epoch": 2.1195180722891567,
"grad_norm": 0.04400300234556198,
"learning_rate": 9.691567537832964e-06,
"loss": 0.0031,
"step": 551
},
{
"epoch": 2.1233734939759037,
"grad_norm": 0.047807373106479645,
"learning_rate": 9.614971699809081e-06,
"loss": 0.0055,
"step": 552
},
{
"epoch": 2.1272289156626507,
"grad_norm": 0.10606139898300171,
"learning_rate": 9.538583847008452e-06,
"loss": 0.0106,
"step": 553
},
{
"epoch": 2.1310843373493977,
"grad_norm": 0.03324699401855469,
"learning_rate": 9.462405509281937e-06,
"loss": 0.0016,
"step": 554
},
{
"epoch": 2.1349397590361447,
"grad_norm": 0.06505841016769409,
"learning_rate": 9.386438212284372e-06,
"loss": 0.0051,
"step": 555
},
{
"epoch": 2.1387951807228918,
"grad_norm": 0.04434973746538162,
"learning_rate": 9.310683477443979e-06,
"loss": 0.004,
"step": 556
},
{
"epoch": 2.1426506024096383,
"grad_norm": 0.0856984406709671,
"learning_rate": 9.235142821931928e-06,
"loss": 0.0054,
"step": 557
},
{
"epoch": 2.1465060240963854,
"grad_norm": 0.11251796036958694,
"learning_rate": 9.159817758631923e-06,
"loss": 0.0098,
"step": 558
},
{
"epoch": 2.1503614457831324,
"grad_norm": 0.036301273852586746,
"learning_rate": 9.084709796109907e-06,
"loss": 0.0024,
"step": 559
},
{
"epoch": 2.1542168674698794,
"grad_norm": 0.03839263319969177,
"learning_rate": 9.009820438583881e-06,
"loss": 0.0017,
"step": 560
},
{
"epoch": 2.1580722891566264,
"grad_norm": 0.8053962588310242,
"learning_rate": 8.93515118589373e-06,
"loss": 0.0037,
"step": 561
},
{
"epoch": 2.1619277108433734,
"grad_norm": 0.05949236825108528,
"learning_rate": 8.860703533471218e-06,
"loss": 0.0067,
"step": 562
},
{
"epoch": 2.1657831325301204,
"grad_norm": 0.03886352851986885,
"learning_rate": 8.786478972310023e-06,
"loss": 0.0021,
"step": 563
},
{
"epoch": 2.1696385542168675,
"grad_norm": 0.0781722217798233,
"learning_rate": 8.712478988935904e-06,
"loss": 0.002,
"step": 564
},
{
"epoch": 2.1734939759036145,
"grad_norm": 0.04123575612902641,
"learning_rate": 8.638705065376887e-06,
"loss": 0.0016,
"step": 565
},
{
"epoch": 2.1773493975903615,
"grad_norm": 0.04534149542450905,
"learning_rate": 8.565158679133605e-06,
"loss": 0.0019,
"step": 566
},
{
"epoch": 2.1812048192771085,
"grad_norm": 0.0474507175385952,
"learning_rate": 8.491841303149728e-06,
"loss": 0.0052,
"step": 567
},
{
"epoch": 2.1850602409638555,
"grad_norm": 0.040683846920728683,
"learning_rate": 8.418754405782423e-06,
"loss": 0.0017,
"step": 568
},
{
"epoch": 2.1889156626506026,
"grad_norm": 0.13555528223514557,
"learning_rate": 8.345899450772975e-06,
"loss": 0.0047,
"step": 569
},
{
"epoch": 2.1927710843373496,
"grad_norm": 0.0471130795776844,
"learning_rate": 8.27327789721746e-06,
"loss": 0.0015,
"step": 570
},
{
"epoch": 2.1966265060240966,
"grad_norm": 0.07882878184318542,
"learning_rate": 8.200891199537549e-06,
"loss": 0.0015,
"step": 571
},
{
"epoch": 2.200481927710843,
"grad_norm": 0.07605530321598053,
"learning_rate": 8.128740807451311e-06,
"loss": 0.0025,
"step": 572
},
{
"epoch": 2.20433734939759,
"grad_norm": 0.0734749510884285,
"learning_rate": 8.056828165944282e-06,
"loss": 0.0111,
"step": 573
},
{
"epoch": 2.208192771084337,
"grad_norm": 0.07274612039327621,
"learning_rate": 7.98515471524042e-06,
"loss": 0.0036,
"step": 574
},
{
"epoch": 2.212048192771084,
"grad_norm": 0.1987583339214325,
"learning_rate": 7.913721890773354e-06,
"loss": 0.0056,
"step": 575
},
{
"epoch": 2.2159036144578312,
"grad_norm": 0.047146931290626526,
"learning_rate": 7.842531123157562e-06,
"loss": 0.0022,
"step": 576
},
{
"epoch": 2.2197590361445783,
"grad_norm": 0.11681365221738815,
"learning_rate": 7.771583838159756e-06,
"loss": 0.0065,
"step": 577
},
{
"epoch": 2.2236144578313253,
"grad_norm": 0.04616257920861244,
"learning_rate": 7.700881456670342e-06,
"loss": 0.003,
"step": 578
},
{
"epoch": 2.2274698795180723,
"grad_norm": 0.06335612386465073,
"learning_rate": 7.630425394674903e-06,
"loss": 0.0047,
"step": 579
},
{
"epoch": 2.2313253012048193,
"grad_norm": 0.051160894334316254,
"learning_rate": 7.560217063225923e-06,
"loss": 0.0083,
"step": 580
},
{
"epoch": 2.2351807228915663,
"grad_norm": 0.0413651168346405,
"learning_rate": 7.49025786841445e-06,
"loss": 0.0017,
"step": 581
},
{
"epoch": 2.2390361445783133,
"grad_norm": 0.04799462854862213,
"learning_rate": 7.420549211342005e-06,
"loss": 0.007,
"step": 582
},
{
"epoch": 2.2428915662650604,
"grad_norm": 2.911201000213623,
"learning_rate": 7.3510924880924575e-06,
"loss": 0.0161,
"step": 583
},
{
"epoch": 2.2467469879518074,
"grad_norm": 0.22194476425647736,
"learning_rate": 7.2818890897041065e-06,
"loss": 0.0044,
"step": 584
},
{
"epoch": 2.2506024096385544,
"grad_norm": 0.04791396111249924,
"learning_rate": 7.212940402141808e-06,
"loss": 0.0026,
"step": 585
},
{
"epoch": 2.2544578313253014,
"grad_norm": 0.042130209505558014,
"learning_rate": 7.1442478062692135e-06,
"loss": 0.0074,
"step": 586
},
{
"epoch": 2.258313253012048,
"grad_norm": 0.056368812918663025,
"learning_rate": 7.075812677821145e-06,
"loss": 0.002,
"step": 587
},
{
"epoch": 2.262168674698795,
"grad_norm": 0.032533109188079834,
"learning_rate": 7.0076363873759865e-06,
"loss": 0.002,
"step": 588
},
{
"epoch": 2.266024096385542,
"grad_norm": 0.025718998163938522,
"learning_rate": 6.939720300328303e-06,
"loss": 0.002,
"step": 589
},
{
"epoch": 2.269879518072289,
"grad_norm": 0.05983180180191994,
"learning_rate": 6.8720657768614115e-06,
"loss": 0.0065,
"step": 590
},
{
"epoch": 2.273734939759036,
"grad_norm": 0.16601082682609558,
"learning_rate": 6.8046741719202385e-06,
"loss": 0.0064,
"step": 591
},
{
"epoch": 2.277590361445783,
"grad_norm": 0.05689080432057381,
"learning_rate": 6.737546835184101e-06,
"loss": 0.0036,
"step": 592
},
{
"epoch": 2.28144578313253,
"grad_norm": 0.050703030079603195,
"learning_rate": 6.67068511103971e-06,
"loss": 0.0063,
"step": 593
},
{
"epoch": 2.285301204819277,
"grad_norm": 0.08983147144317627,
"learning_rate": 6.604090338554267e-06,
"loss": 0.003,
"step": 594
},
{
"epoch": 2.289156626506024,
"grad_norm": 0.045743219554424286,
"learning_rate": 6.537763851448593e-06,
"loss": 0.0033,
"step": 595
},
{
"epoch": 2.293012048192771,
"grad_norm": 0.034684326499700546,
"learning_rate": 6.471706978070482e-06,
"loss": 0.0019,
"step": 596
},
{
"epoch": 2.296867469879518,
"grad_norm": 0.04284167289733887,
"learning_rate": 6.4059210413680175e-06,
"loss": 0.0021,
"step": 597
},
{
"epoch": 2.300722891566265,
"grad_norm": 0.05777349695563316,
"learning_rate": 6.340407358863167e-06,
"loss": 0.0032,
"step": 598
},
{
"epoch": 2.304578313253012,
"grad_norm": 0.058219652622938156,
"learning_rate": 6.275167242625331e-06,
"loss": 0.0049,
"step": 599
},
{
"epoch": 2.3084337349397592,
"grad_norm": 0.06586357206106186,
"learning_rate": 6.210201999245087e-06,
"loss": 0.0045,
"step": 600
},
{
"epoch": 2.3122891566265062,
"grad_norm": 0.0502832792699337,
"learning_rate": 6.145512929808013e-06,
"loss": 0.0028,
"step": 601
},
{
"epoch": 2.316144578313253,
"grad_norm": 0.051489003002643585,
"learning_rate": 6.081101329868664e-06,
"loss": 0.0065,
"step": 602
},
{
"epoch": 2.32,
"grad_norm": 0.07483097910881042,
"learning_rate": 6.016968489424572e-06,
"loss": 0.0069,
"step": 603
},
{
"epoch": 2.323855421686747,
"grad_norm": 0.0629890039563179,
"learning_rate": 5.95311569289045e-06,
"loss": 0.0016,
"step": 604
},
{
"epoch": 2.327710843373494,
"grad_norm": 0.04719253256917,
"learning_rate": 5.889544219072465e-06,
"loss": 0.0016,
"step": 605
},
{
"epoch": 2.331566265060241,
"grad_norm": 0.04632150009274483,
"learning_rate": 5.8262553411426015e-06,
"loss": 0.0024,
"step": 606
},
{
"epoch": 2.335421686746988,
"grad_norm": 0.03981681540608406,
"learning_rate": 5.7632503266131925e-06,
"loss": 0.0059,
"step": 607
},
{
"epoch": 2.339277108433735,
"grad_norm": 0.032810911536216736,
"learning_rate": 5.700530437311509e-06,
"loss": 0.0017,
"step": 608
},
{
"epoch": 2.343132530120482,
"grad_norm": 0.252675324678421,
"learning_rate": 5.638096929354522e-06,
"loss": 0.0031,
"step": 609
},
{
"epoch": 2.346987951807229,
"grad_norm": 0.09215971827507019,
"learning_rate": 5.5759510531237116e-06,
"loss": 0.0196,
"step": 610
},
{
"epoch": 2.350843373493976,
"grad_norm": 0.05282469466328621,
"learning_rate": 5.514094053240035e-06,
"loss": 0.0033,
"step": 611
},
{
"epoch": 2.354698795180723,
"grad_norm": 0.04628681764006615,
"learning_rate": 5.452527168539026e-06,
"loss": 0.0016,
"step": 612
},
{
"epoch": 2.35855421686747,
"grad_norm": 0.05543181672692299,
"learning_rate": 5.39125163204594e-06,
"loss": 0.0051,
"step": 613
},
{
"epoch": 2.362409638554217,
"grad_norm": 0.04562715068459511,
"learning_rate": 5.330268670951091e-06,
"loss": 0.0015,
"step": 614
},
{
"epoch": 2.3662650602409636,
"grad_norm": 0.05793346092104912,
"learning_rate": 5.269579506585259e-06,
"loss": 0.0069,
"step": 615
},
{
"epoch": 2.370120481927711,
"grad_norm": 0.06059642881155014,
"learning_rate": 5.209185354395261e-06,
"loss": 0.0029,
"step": 616
},
{
"epoch": 2.3739759036144576,
"grad_norm": 0.04422873258590698,
"learning_rate": 5.149087423919541e-06,
"loss": 0.0028,
"step": 617
},
{
"epoch": 2.3778313253012047,
"grad_norm": 0.088904470205307,
"learning_rate": 5.089286918764031e-06,
"loss": 0.0036,
"step": 618
},
{
"epoch": 2.3816867469879517,
"grad_norm": 0.05677289515733719,
"learning_rate": 5.029785036577976e-06,
"loss": 0.0043,
"step": 619
},
{
"epoch": 2.3855421686746987,
"grad_norm": 0.08604881912469864,
"learning_rate": 4.97058296903e-06,
"loss": 0.0061,
"step": 620
},
{
"epoch": 2.3893975903614457,
"grad_norm": 0.037634607404470444,
"learning_rate": 4.911681901784198e-06,
"loss": 0.0025,
"step": 621
},
{
"epoch": 2.3932530120481927,
"grad_norm": 0.05942295491695404,
"learning_rate": 4.853083014476403e-06,
"loss": 0.0035,
"step": 622
},
{
"epoch": 2.3971084337349398,
"grad_norm": 0.050786908715963364,
"learning_rate": 4.794787480690597e-06,
"loss": 0.0042,
"step": 623
},
{
"epoch": 2.4009638554216868,
"grad_norm": 0.19012530148029327,
"learning_rate": 4.736796467935336e-06,
"loss": 0.0237,
"step": 624
},
{
"epoch": 2.404819277108434,
"grad_norm": 0.07656612247228622,
"learning_rate": 4.679111137620442e-06,
"loss": 0.0043,
"step": 625
},
{
"epoch": 2.408674698795181,
"grad_norm": 0.05607719346880913,
"learning_rate": 4.621732645033687e-06,
"loss": 0.0014,
"step": 626
},
{
"epoch": 2.412530120481928,
"grad_norm": 0.042300060391426086,
"learning_rate": 4.5646621393177e-06,
"loss": 0.0013,
"step": 627
},
{
"epoch": 2.416385542168675,
"grad_norm": 0.07272662222385406,
"learning_rate": 4.507900763446911e-06,
"loss": 0.0046,
"step": 628
},
{
"epoch": 2.420240963855422,
"grad_norm": 0.1015816405415535,
"learning_rate": 4.451449654204685e-06,
"loss": 0.0025,
"step": 629
},
{
"epoch": 2.4240963855421684,
"grad_norm": 0.08338016271591187,
"learning_rate": 4.395309942160551e-06,
"loss": 0.0044,
"step": 630
},
{
"epoch": 2.427951807228916,
"grad_norm": 0.04506392404437065,
"learning_rate": 4.339482751647557e-06,
"loss": 0.0021,
"step": 631
},
{
"epoch": 2.4318072289156625,
"grad_norm": 0.04617151618003845,
"learning_rate": 4.283969200739764e-06,
"loss": 0.0042,
"step": 632
},
{
"epoch": 2.4356626506024095,
"grad_norm": 0.04533221572637558,
"learning_rate": 4.228770401229824e-06,
"loss": 0.0021,
"step": 633
},
{
"epoch": 2.4395180722891565,
"grad_norm": 0.03105015493929386,
"learning_rate": 4.173887458606767e-06,
"loss": 0.0015,
"step": 634
},
{
"epoch": 2.4433734939759035,
"grad_norm": 0.03597990423440933,
"learning_rate": 4.119321472033779e-06,
"loss": 0.0031,
"step": 635
},
{
"epoch": 2.4472289156626506,
"grad_norm": 0.036103278398513794,
"learning_rate": 4.065073534326282e-06,
"loss": 0.0024,
"step": 636
},
{
"epoch": 2.4510843373493976,
"grad_norm": 0.09541872888803482,
"learning_rate": 4.011144731929981e-06,
"loss": 0.0146,
"step": 637
},
{
"epoch": 2.4549397590361446,
"grad_norm": 0.05025950074195862,
"learning_rate": 3.957536144899123e-06,
"loss": 0.0027,
"step": 638
},
{
"epoch": 2.4587951807228916,
"grad_norm": 0.08015301078557968,
"learning_rate": 3.904248846874894e-06,
"loss": 0.0189,
"step": 639
},
{
"epoch": 2.4626506024096386,
"grad_norm": 0.042166490107774734,
"learning_rate": 3.851283905063863e-06,
"loss": 0.0027,
"step": 640
},
{
"epoch": 2.4665060240963856,
"grad_norm": 0.0753423422574997,
"learning_rate": 3.7986423802166705e-06,
"loss": 0.0111,
"step": 641
},
{
"epoch": 2.4703614457831327,
"grad_norm": 0.09548261016607285,
"learning_rate": 3.7463253266067147e-06,
"loss": 0.0092,
"step": 642
},
{
"epoch": 2.4742168674698797,
"grad_norm": 0.03466320037841797,
"learning_rate": 3.694333792009115e-06,
"loss": 0.0015,
"step": 643
},
{
"epoch": 2.4780722891566267,
"grad_norm": 0.03481588885188103,
"learning_rate": 3.64266881767966e-06,
"loss": 0.0013,
"step": 644
},
{
"epoch": 2.4819277108433733,
"grad_norm": 0.07214005291461945,
"learning_rate": 3.5913314383339937e-06,
"loss": 0.002,
"step": 645
},
{
"epoch": 2.4857831325301207,
"grad_norm": 0.03147978335618973,
"learning_rate": 3.5403226821268734e-06,
"loss": 0.0014,
"step": 646
},
{
"epoch": 2.4896385542168673,
"grad_norm": 0.08119364082813263,
"learning_rate": 3.4896435706316e-06,
"loss": 0.0034,
"step": 647
},
{
"epoch": 2.4934939759036143,
"grad_norm": 0.09093235433101654,
"learning_rate": 3.43929511881953e-06,
"loss": 0.0064,
"step": 648
},
{
"epoch": 2.4973493975903613,
"grad_norm": 0.04797382652759552,
"learning_rate": 3.3892783350397675e-06,
"loss": 0.0028,
"step": 649
},
{
"epoch": 2.5012048192771084,
"grad_norm": 0.03735937178134918,
"learning_rate": 3.3395942209989694e-06,
"loss": 0.0012,
"step": 650
},
{
"epoch": 2.5050602409638554,
"grad_norm": 0.0818742886185646,
"learning_rate": 3.290243771741275e-06,
"loss": 0.0103,
"step": 651
},
{
"epoch": 2.5089156626506024,
"grad_norm": 0.06941164284944534,
"learning_rate": 3.241227975628378e-06,
"loss": 0.0037,
"step": 652
},
{
"epoch": 2.5127710843373494,
"grad_norm": 0.0699724406003952,
"learning_rate": 3.1925478143197418e-06,
"loss": 0.0048,
"step": 653
},
{
"epoch": 2.5166265060240964,
"grad_norm": 0.04508174955844879,
"learning_rate": 3.144204262752937e-06,
"loss": 0.0024,
"step": 654
},
{
"epoch": 2.5204819277108435,
"grad_norm": 0.05046376213431358,
"learning_rate": 3.0961982891241083e-06,
"loss": 0.0051,
"step": 655
},
{
"epoch": 2.5243373493975905,
"grad_norm": 0.058411385864019394,
"learning_rate": 3.048530854868579e-06,
"loss": 0.0032,
"step": 656
},
{
"epoch": 2.5281927710843375,
"grad_norm": 0.04653112217783928,
"learning_rate": 3.001202914641628e-06,
"loss": 0.0021,
"step": 657
},
{
"epoch": 2.532048192771084,
"grad_norm": 0.08640334010124207,
"learning_rate": 2.954215416299331e-06,
"loss": 0.0069,
"step": 658
},
{
"epoch": 2.5359036144578315,
"grad_norm": 0.06023968756198883,
"learning_rate": 2.907569300879596e-06,
"loss": 0.0054,
"step": 659
},
{
"epoch": 2.539759036144578,
"grad_norm": 0.04500487819314003,
"learning_rate": 2.8612655025833148e-06,
"loss": 0.0038,
"step": 660
},
{
"epoch": 2.5436144578313256,
"grad_norm": 0.20122352242469788,
"learning_rate": 2.815304948755664e-06,
"loss": 0.0049,
"step": 661
},
{
"epoch": 2.547469879518072,
"grad_norm": 0.055998221039772034,
"learning_rate": 2.769688559867514e-06,
"loss": 0.0051,
"step": 662
},
{
"epoch": 2.551325301204819,
"grad_norm": 0.08072318136692047,
"learning_rate": 2.7244172494969978e-06,
"loss": 0.0084,
"step": 663
},
{
"epoch": 2.555180722891566,
"grad_norm": 0.036920543760061264,
"learning_rate": 2.679491924311226e-06,
"loss": 0.0015,
"step": 664
},
{
"epoch": 2.559036144578313,
"grad_norm": 0.08001082390546799,
"learning_rate": 2.6349134840481294e-06,
"loss": 0.0039,
"step": 665
},
{
"epoch": 2.56289156626506,
"grad_norm": 0.028426064178347588,
"learning_rate": 2.590682821498416e-06,
"loss": 0.0019,
"step": 666
},
{
"epoch": 2.5667469879518072,
"grad_norm": 0.061366964131593704,
"learning_rate": 2.546800822487714e-06,
"loss": 0.0035,
"step": 667
},
{
"epoch": 2.5706024096385542,
"grad_norm": 0.050951141864061356,
"learning_rate": 2.503268365858831e-06,
"loss": 0.0031,
"step": 668
},
{
"epoch": 2.5744578313253013,
"grad_norm": 0.053869444876909256,
"learning_rate": 2.4600863234541338e-06,
"loss": 0.0023,
"step": 669
},
{
"epoch": 2.5783132530120483,
"grad_norm": 0.042749661952257156,
"learning_rate": 2.417255560098111e-06,
"loss": 0.0017,
"step": 670
},
{
"epoch": 2.5821686746987953,
"grad_norm": 0.06091522425413132,
"learning_rate": 2.374776933580025e-06,
"loss": 0.0043,
"step": 671
},
{
"epoch": 2.5860240963855423,
"grad_norm": 0.045107800513505936,
"learning_rate": 2.3326512946367717e-06,
"loss": 0.0019,
"step": 672
},
{
"epoch": 2.589879518072289,
"grad_norm": 0.05708494782447815,
"learning_rate": 2.2908794869358044e-06,
"loss": 0.0104,
"step": 673
},
{
"epoch": 2.5937349397590364,
"grad_norm": 0.04535680264234543,
"learning_rate": 2.249462347058251e-06,
"loss": 0.0031,
"step": 674
},
{
"epoch": 2.597590361445783,
"grad_norm": 0.04893667623400688,
"learning_rate": 2.2084007044821764e-06,
"loss": 0.0057,
"step": 675
},
{
"epoch": 2.6014457831325304,
"grad_norm": 0.04653600975871086,
"learning_rate": 2.167695381565944e-06,
"loss": 0.0038,
"step": 676
},
{
"epoch": 2.605301204819277,
"grad_norm": 0.05264260247349739,
"learning_rate": 2.127347193531757e-06,
"loss": 0.0059,
"step": 677
},
{
"epoch": 2.609156626506024,
"grad_norm": 0.04379219189286232,
"learning_rate": 2.0873569484493305e-06,
"loss": 0.0031,
"step": 678
},
{
"epoch": 2.613012048192771,
"grad_norm": 0.07054059207439423,
"learning_rate": 2.0477254472197237e-06,
"loss": 0.0072,
"step": 679
},
{
"epoch": 2.616867469879518,
"grad_norm": 0.06054523587226868,
"learning_rate": 2.0084534835592584e-06,
"loss": 0.0062,
"step": 680
},
{
"epoch": 2.620722891566265,
"grad_norm": 0.0372784286737442,
"learning_rate": 1.96954184398368e-06,
"loss": 0.0018,
"step": 681
},
{
"epoch": 2.624578313253012,
"grad_norm": 0.05901319161057472,
"learning_rate": 1.930991307792358e-06,
"loss": 0.0023,
"step": 682
},
{
"epoch": 2.628433734939759,
"grad_norm": 0.059202998876571655,
"learning_rate": 1.8928026470526917e-06,
"loss": 0.0037,
"step": 683
},
{
"epoch": 2.632289156626506,
"grad_norm": 0.04968271404504776,
"learning_rate": 1.8549766265846725e-06,
"loss": 0.0027,
"step": 684
},
{
"epoch": 2.636144578313253,
"grad_norm": 0.06249796226620674,
"learning_rate": 1.817514003945524e-06,
"loss": 0.01,
"step": 685
},
{
"epoch": 2.64,
"grad_norm": 0.03794616460800171,
"learning_rate": 1.7804155294145765e-06,
"loss": 0.0014,
"step": 686
},
{
"epoch": 2.643855421686747,
"grad_norm": 0.032206736505031586,
"learning_rate": 1.743681945978184e-06,
"loss": 0.0024,
"step": 687
},
{
"epoch": 2.6477108433734937,
"grad_norm": 0.09019238501787186,
"learning_rate": 1.7073139893149092e-06,
"loss": 0.0117,
"step": 688
},
{
"epoch": 2.651566265060241,
"grad_norm": 0.03745673596858978,
"learning_rate": 1.6713123877807413e-06,
"loss": 0.0028,
"step": 689
},
{
"epoch": 2.6554216867469878,
"grad_norm": 0.05633678659796715,
"learning_rate": 1.6356778623945223e-06,
"loss": 0.0039,
"step": 690
},
{
"epoch": 2.659277108433735,
"grad_norm": 0.06637841463088989,
"learning_rate": 1.6004111268235156e-06,
"loss": 0.0049,
"step": 691
},
{
"epoch": 2.663132530120482,
"grad_norm": 0.03530237078666687,
"learning_rate": 1.5655128873691138e-06,
"loss": 0.0015,
"step": 692
},
{
"epoch": 2.666987951807229,
"grad_norm": 0.043858692049980164,
"learning_rate": 1.5309838429526714e-06,
"loss": 0.0031,
"step": 693
},
{
"epoch": 2.670843373493976,
"grad_norm": 0.0662481039762497,
"learning_rate": 1.49682468510153e-06,
"loss": 0.0049,
"step": 694
},
{
"epoch": 2.674698795180723,
"grad_norm": 0.04087715223431587,
"learning_rate": 1.4630360979351644e-06,
"loss": 0.0017,
"step": 695
},
{
"epoch": 2.67855421686747,
"grad_norm": 0.04193538799881935,
"learning_rate": 1.4296187581514742e-06,
"loss": 0.0049,
"step": 696
},
{
"epoch": 2.682409638554217,
"grad_norm": 0.045852597802877426,
"learning_rate": 1.396573335013236e-06,
"loss": 0.0018,
"step": 697
},
{
"epoch": 2.686265060240964,
"grad_norm": 0.06638387590646744,
"learning_rate": 1.3639004903346954e-06,
"loss": 0.0097,
"step": 698
},
{
"epoch": 2.690120481927711,
"grad_norm": 1.4415384531021118,
"learning_rate": 1.3316008784683265e-06,
"loss": 0.0567,
"step": 699
},
{
"epoch": 2.693975903614458,
"grad_norm": 0.03742547705769539,
"learning_rate": 1.2996751462917057e-06,
"loss": 0.0013,
"step": 700
},
{
"epoch": 2.697831325301205,
"grad_norm": 0.053348954766988754,
"learning_rate": 1.2681239331945695e-06,
"loss": 0.003,
"step": 701
},
{
"epoch": 2.701686746987952,
"grad_norm": 0.054243747144937515,
"learning_rate": 1.2369478710660166e-06,
"loss": 0.0032,
"step": 702
},
{
"epoch": 2.7055421686746985,
"grad_norm": 0.03384071961045265,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.001,
"step": 703
},
{
"epoch": 2.709397590361446,
"grad_norm": 0.06644614785909653,
"learning_rate": 1.175723689692001e-06,
"loss": 0.0087,
"step": 704
},
{
"epoch": 2.7132530120481926,
"grad_norm": 0.03085707314312458,
"learning_rate": 1.1456767966083393e-06,
"loss": 0.0013,
"step": 705
},
{
"epoch": 2.7171084337349396,
"grad_norm": 0.04153566062450409,
"learning_rate": 1.1160075067923092e-06,
"loss": 0.0032,
"step": 706
},
{
"epoch": 2.7209638554216866,
"grad_norm": 0.03273540362715721,
"learning_rate": 1.086716414442952e-06,
"loss": 0.0009,
"step": 707
},
{
"epoch": 2.7248192771084336,
"grad_norm": 0.06003859266638756,
"learning_rate": 1.057804106184992e-06,
"loss": 0.0021,
"step": 708
},
{
"epoch": 2.7286746987951807,
"grad_norm": 0.024825889617204666,
"learning_rate": 1.0292711610570904e-06,
"loss": 0.0012,
"step": 709
},
{
"epoch": 2.7325301204819277,
"grad_norm": 0.04767761379480362,
"learning_rate": 1.0011181505002488e-06,
"loss": 0.0022,
"step": 710
},
{
"epoch": 2.7363855421686747,
"grad_norm": 0.07471596449613571,
"learning_rate": 9.733456383463658e-07,
"loss": 0.002,
"step": 711
},
{
"epoch": 2.7402409638554217,
"grad_norm": 0.02973441407084465,
"learning_rate": 9.459541808069295e-07,
"loss": 0.0011,
"step": 712
},
{
"epoch": 2.7440963855421687,
"grad_norm": 0.057284433394670486,
"learning_rate": 9.189443264619102e-07,
"loss": 0.0029,
"step": 713
},
{
"epoch": 2.7479518072289157,
"grad_norm": 0.040791746228933334,
"learning_rate": 8.923166162487407e-07,
"loss": 0.0036,
"step": 714
},
{
"epoch": 2.7518072289156628,
"grad_norm": 0.04032406583428383,
"learning_rate": 8.660715834514977e-07,
"loss": 0.0034,
"step": 715
},
{
"epoch": 2.75566265060241,
"grad_norm": 0.16751016676425934,
"learning_rate": 8.402097536902221e-07,
"loss": 0.004,
"step": 716
},
{
"epoch": 2.759518072289157,
"grad_norm": 0.047715965658426285,
"learning_rate": 8.147316449103959e-07,
"loss": 0.0027,
"step": 717
},
{
"epoch": 2.7633734939759034,
"grad_norm": 0.057923998683691025,
"learning_rate": 7.896377673725553e-07,
"loss": 0.0042,
"step": 718
},
{
"epoch": 2.767228915662651,
"grad_norm": 0.1632152944803238,
"learning_rate": 7.649286236420806e-07,
"loss": 0.0062,
"step": 719
},
{
"epoch": 2.7710843373493974,
"grad_norm": 0.03497255966067314,
"learning_rate": 7.406047085791379e-07,
"loss": 0.0014,
"step": 720
},
{
"epoch": 2.7749397590361444,
"grad_norm": 0.05902968347072601,
"learning_rate": 7.166665093287539e-07,
"loss": 0.0054,
"step": 721
},
{
"epoch": 2.7787951807228914,
"grad_norm": 0.04337353631854057,
"learning_rate": 6.931145053110655e-07,
"loss": 0.0046,
"step": 722
},
{
"epoch": 2.7826506024096385,
"grad_norm": 0.057529669255018234,
"learning_rate": 6.69949168211721e-07,
"loss": 0.006,
"step": 723
},
{
"epoch": 2.7865060240963855,
"grad_norm": 0.03574662655591965,
"learning_rate": 6.471709619724386e-07,
"loss": 0.0038,
"step": 724
},
{
"epoch": 2.7903614457831325,
"grad_norm": 0.05454541742801666,
"learning_rate": 6.247803427816945e-07,
"loss": 0.0034,
"step": 725
},
{
"epoch": 2.7942168674698795,
"grad_norm": 0.05086322873830795,
"learning_rate": 6.027777590656115e-07,
"loss": 0.0049,
"step": 726
},
{
"epoch": 2.7980722891566265,
"grad_norm": 0.06615273654460907,
"learning_rate": 5.811636514789598e-07,
"loss": 0.0016,
"step": 727
},
{
"epoch": 2.8019277108433736,
"grad_norm": 0.05431687459349632,
"learning_rate": 5.599384528963425e-07,
"loss": 0.0054,
"step": 728
},
{
"epoch": 2.8057831325301206,
"grad_norm": 0.03179425001144409,
"learning_rate": 5.391025884035239e-07,
"loss": 0.0021,
"step": 729
},
{
"epoch": 2.8096385542168676,
"grad_norm": 0.048835258930921555,
"learning_rate": 5.186564752889078e-07,
"loss": 0.0044,
"step": 730
},
{
"epoch": 2.8134939759036146,
"grad_norm": 0.052003178745508194,
"learning_rate": 4.986005230351954e-07,
"loss": 0.0022,
"step": 731
},
{
"epoch": 2.8173493975903616,
"grad_norm": 0.04256909340620041,
"learning_rate": 4.789351333111691e-07,
"loss": 0.0017,
"step": 732
},
{
"epoch": 2.821204819277108,
"grad_norm": 0.03870426118373871,
"learning_rate": 4.5966069996365993e-07,
"loss": 0.0038,
"step": 733
},
{
"epoch": 2.8250602409638557,
"grad_norm": 0.08713089674711227,
"learning_rate": 4.407776090096527e-07,
"loss": 0.0102,
"step": 734
},
{
"epoch": 2.8289156626506022,
"grad_norm": 0.0968024879693985,
"learning_rate": 4.22286238628562e-07,
"loss": 0.0041,
"step": 735
},
{
"epoch": 2.8327710843373493,
"grad_norm": 0.040818337351083755,
"learning_rate": 4.0418695915464657e-07,
"loss": 0.0049,
"step": 736
},
{
"epoch": 2.8366265060240963,
"grad_norm": 0.0483890138566494,
"learning_rate": 3.8648013306960664e-07,
"loss": 0.0077,
"step": 737
},
{
"epoch": 2.8404819277108433,
"grad_norm": 0.03612169250845909,
"learning_rate": 3.691661149953096e-07,
"loss": 0.0017,
"step": 738
},
{
"epoch": 2.8443373493975903,
"grad_norm": 0.058921512216329575,
"learning_rate": 3.522452516867048e-07,
"loss": 0.003,
"step": 739
},
{
"epoch": 2.8481927710843373,
"grad_norm": 0.04771070182323456,
"learning_rate": 3.357178820248641e-07,
"loss": 0.0031,
"step": 740
},
{
"epoch": 2.8520481927710843,
"grad_norm": 0.04708501696586609,
"learning_rate": 3.1958433701019697e-07,
"loss": 0.0031,
"step": 741
},
{
"epoch": 2.8559036144578314,
"grad_norm": 0.05502479150891304,
"learning_rate": 3.038449397558396e-07,
"loss": 0.0103,
"step": 742
},
{
"epoch": 2.8597590361445784,
"grad_norm": 0.063191719353199,
"learning_rate": 2.8850000548115155e-07,
"loss": 0.0089,
"step": 743
},
{
"epoch": 2.8636144578313254,
"grad_norm": 0.047001924365758896,
"learning_rate": 2.735498415054294e-07,
"loss": 0.0021,
"step": 744
},
{
"epoch": 2.8674698795180724,
"grad_norm": 0.07529550790786743,
"learning_rate": 2.5899474724174313e-07,
"loss": 0.0101,
"step": 745
},
{
"epoch": 2.8713253012048194,
"grad_norm": 0.05167806148529053,
"learning_rate": 2.4483501419093836e-07,
"loss": 0.0073,
"step": 746
},
{
"epoch": 2.8751807228915665,
"grad_norm": 0.20707423985004425,
"learning_rate": 2.3107092593579905e-07,
"loss": 0.0071,
"step": 747
},
{
"epoch": 2.879036144578313,
"grad_norm": 0.04456199333071709,
"learning_rate": 2.1770275813536746e-07,
"loss": 0.004,
"step": 748
},
{
"epoch": 2.8828915662650605,
"grad_norm": 0.0451984740793705,
"learning_rate": 2.0473077851942858e-07,
"loss": 0.003,
"step": 749
},
{
"epoch": 2.886746987951807,
"grad_norm": 0.1236504539847374,
"learning_rate": 1.9215524688313893e-07,
"loss": 0.0124,
"step": 750
},
{
"epoch": 2.890602409638554,
"grad_norm": 0.037482503801584244,
"learning_rate": 1.799764150818306e-07,
"loss": 0.002,
"step": 751
},
{
"epoch": 2.894457831325301,
"grad_norm": 0.04179581627249718,
"learning_rate": 1.6819452702596882e-07,
"loss": 0.0033,
"step": 752
},
{
"epoch": 2.898313253012048,
"grad_norm": 0.03997666761279106,
"learning_rate": 1.5680981867625566e-07,
"loss": 0.0015,
"step": 753
},
{
"epoch": 2.902168674698795,
"grad_norm": 0.03578900918364525,
"learning_rate": 1.4582251803892055e-07,
"loss": 0.002,
"step": 754
},
{
"epoch": 2.906024096385542,
"grad_norm": 0.056817784905433655,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.0132,
"step": 755
},
{
"epoch": 2.909879518072289,
"grad_norm": 0.06569056212902069,
"learning_rate": 1.2504101212664099e-07,
"loss": 0.0085,
"step": 756
},
{
"epoch": 2.913734939759036,
"grad_norm": 0.04236573725938797,
"learning_rate": 1.1524722305144231e-07,
"loss": 0.0031,
"step": 757
},
{
"epoch": 2.917590361445783,
"grad_norm": 0.03456762433052063,
"learning_rate": 1.058516740797777e-07,
"loss": 0.0013,
"step": 758
},
{
"epoch": 2.9214457831325302,
"grad_norm": 0.05244007334113121,
"learning_rate": 9.685455338016347e-08,
"loss": 0.0022,
"step": 759
},
{
"epoch": 2.9253012048192772,
"grad_norm": 0.06270337104797363,
"learning_rate": 8.825604114162556e-08,
"loss": 0.0038,
"step": 760
},
{
"epoch": 2.929156626506024,
"grad_norm": 0.05413990467786789,
"learning_rate": 8.005630957010014e-08,
"loss": 0.0033,
"step": 761
},
{
"epoch": 2.9330120481927713,
"grad_norm": 0.07030560076236725,
"learning_rate": 7.225552288497195e-08,
"loss": 0.0043,
"step": 762
},
{
"epoch": 2.936867469879518,
"grad_norm": 0.07556058466434479,
"learning_rate": 6.485383731580142e-08,
"loss": 0.0114,
"step": 763
},
{
"epoch": 2.9407228915662653,
"grad_norm": 0.037275027483701706,
"learning_rate": 5.785140109917819e-08,
"loss": 0.0014,
"step": 764
},
{
"epoch": 2.944578313253012,
"grad_norm": 0.03457137197256088,
"learning_rate": 5.1248354475768034e-08,
"loss": 0.0016,
"step": 765
},
{
"epoch": 2.948433734939759,
"grad_norm": 0.050263967365026474,
"learning_rate": 4.5044829687497236e-08,
"loss": 0.0049,
"step": 766
},
{
"epoch": 2.952289156626506,
"grad_norm": 0.04029664024710655,
"learning_rate": 3.924095097489922e-08,
"loss": 0.0028,
"step": 767
},
{
"epoch": 2.956144578313253,
"grad_norm": 0.03923238068819046,
"learning_rate": 3.383683457463649e-08,
"loss": 0.002,
"step": 768
},
{
"epoch": 2.96,
"grad_norm": 0.034028057008981705,
"learning_rate": 2.8832588717164766e-08,
"loss": 0.0027,
"step": 769
},
{
"epoch": 2.963855421686747,
"grad_norm": 0.12313743680715561,
"learning_rate": 2.4228313624568013e-08,
"loss": 0.0035,
"step": 770
},
{
"epoch": 2.967710843373494,
"grad_norm": 0.098887600004673,
"learning_rate": 2.0024101508555604e-08,
"loss": 0.0104,
"step": 771
},
{
"epoch": 2.971566265060241,
"grad_norm": 0.03440746292471886,
"learning_rate": 1.6220036568608265e-08,
"loss": 0.0013,
"step": 772
},
{
"epoch": 2.975421686746988,
"grad_norm": 0.12951649725437164,
"learning_rate": 1.281619499029274e-08,
"loss": 0.0023,
"step": 773
},
{
"epoch": 2.979277108433735,
"grad_norm": 0.029755258932709694,
"learning_rate": 9.812644943745231e-09,
"loss": 0.0013,
"step": 774
},
{
"epoch": 2.983132530120482,
"grad_norm": 0.0304043497890234,
"learning_rate": 7.209446582292501e-09,
"loss": 0.0012,
"step": 775
},
{
"epoch": 2.9869879518072286,
"grad_norm": 0.04653947055339813,
"learning_rate": 5.006652041255055e-09,
"loss": 0.0022,
"step": 776
},
{
"epoch": 2.990843373493976,
"grad_norm": 0.05108502134680748,
"learning_rate": 3.2043054369057523e-09,
"loss": 0.0016,
"step": 777
},
{
"epoch": 2.9946987951807227,
"grad_norm": 0.13829636573791504,
"learning_rate": 1.8024428655794012e-09,
"loss": 0.0148,
"step": 778
},
{
"epoch": 2.99855421686747,
"grad_norm": 0.05087655037641525,
"learning_rate": 8.010924029533406e-10,
"loss": 0.0101,
"step": 779
},
{
"epoch": 3.0,
"grad_norm": 0.06550304591655731,
"learning_rate": 2.002741034812239e-10,
"loss": 0.0011,
"step": 780
},
{
"epoch": 3.0,
"step": 780,
"total_flos": 6.08604739672526e+18,
"train_loss": 0.010436751544759131,
"train_runtime": 12813.4993,
"train_samples_per_second": 3.885,
"train_steps_per_second": 0.061
}
],
"logging_steps": 1,
"max_steps": 780,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 92,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.08604739672526e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}