2773 lines
63 KiB
JSON
2773 lines
63 KiB
JSON
{
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 4.992,
|
|
"eval_steps": 500,
|
|
"global_step": 390,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.0128,
|
|
"grad_norm": 6.464605218515887,
|
|
"learning_rate": 1.0256410256410257e-06,
|
|
"loss": 0.8325,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.0256,
|
|
"grad_norm": 6.380708677323359,
|
|
"learning_rate": 2.0512820512820513e-06,
|
|
"loss": 0.8289,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.0384,
|
|
"grad_norm": 6.0732844198848,
|
|
"learning_rate": 3.0769230769230774e-06,
|
|
"loss": 0.8078,
|
|
"step": 3
|
|
},
|
|
{
|
|
"epoch": 0.0512,
|
|
"grad_norm": 5.719294948449857,
|
|
"learning_rate": 4.102564102564103e-06,
|
|
"loss": 0.8006,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.064,
|
|
"grad_norm": 4.0447104367675575,
|
|
"learning_rate": 5.128205128205128e-06,
|
|
"loss": 0.7421,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.0768,
|
|
"grad_norm": 2.3143768442428203,
|
|
"learning_rate": 6.153846153846155e-06,
|
|
"loss": 0.7318,
|
|
"step": 6
|
|
},
|
|
{
|
|
"epoch": 0.0896,
|
|
"grad_norm": 2.107989614101519,
|
|
"learning_rate": 7.17948717948718e-06,
|
|
"loss": 0.7258,
|
|
"step": 7
|
|
},
|
|
{
|
|
"epoch": 0.1024,
|
|
"grad_norm": 3.9720378553941766,
|
|
"learning_rate": 8.205128205128205e-06,
|
|
"loss": 0.7241,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.1152,
|
|
"grad_norm": 4.270610484139302,
|
|
"learning_rate": 9.230769230769232e-06,
|
|
"loss": 0.7218,
|
|
"step": 9
|
|
},
|
|
{
|
|
"epoch": 0.128,
|
|
"grad_norm": 4.148966868142642,
|
|
"learning_rate": 1.0256410256410256e-05,
|
|
"loss": 0.6697,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.1408,
|
|
"grad_norm": 3.7021017185511176,
|
|
"learning_rate": 1.1282051282051283e-05,
|
|
"loss": 0.6733,
|
|
"step": 11
|
|
},
|
|
{
|
|
"epoch": 0.1536,
|
|
"grad_norm": 2.142866919387794,
|
|
"learning_rate": 1.230769230769231e-05,
|
|
"loss": 0.6395,
|
|
"step": 12
|
|
},
|
|
{
|
|
"epoch": 0.1664,
|
|
"grad_norm": 1.707078933942607,
|
|
"learning_rate": 1.3333333333333333e-05,
|
|
"loss": 0.6393,
|
|
"step": 13
|
|
},
|
|
{
|
|
"epoch": 0.1792,
|
|
"grad_norm": 2.147594996913022,
|
|
"learning_rate": 1.435897435897436e-05,
|
|
"loss": 0.6228,
|
|
"step": 14
|
|
},
|
|
{
|
|
"epoch": 0.192,
|
|
"grad_norm": 1.5939068186916545,
|
|
"learning_rate": 1.5384615384615387e-05,
|
|
"loss": 0.5948,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.2048,
|
|
"grad_norm": 1.0835969430668775,
|
|
"learning_rate": 1.641025641025641e-05,
|
|
"loss": 0.5971,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.2176,
|
|
"grad_norm": 1.2234307086927172,
|
|
"learning_rate": 1.7435897435897438e-05,
|
|
"loss": 0.5931,
|
|
"step": 17
|
|
},
|
|
{
|
|
"epoch": 0.2304,
|
|
"grad_norm": 0.8928903139357364,
|
|
"learning_rate": 1.8461538461538465e-05,
|
|
"loss": 0.5712,
|
|
"step": 18
|
|
},
|
|
{
|
|
"epoch": 0.2432,
|
|
"grad_norm": 0.9673561701509722,
|
|
"learning_rate": 1.9487179487179488e-05,
|
|
"loss": 0.5674,
|
|
"step": 19
|
|
},
|
|
{
|
|
"epoch": 0.256,
|
|
"grad_norm": 0.8468004286837573,
|
|
"learning_rate": 2.0512820512820512e-05,
|
|
"loss": 0.5393,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.2688,
|
|
"grad_norm": 0.8301118890092068,
|
|
"learning_rate": 2.153846153846154e-05,
|
|
"loss": 0.551,
|
|
"step": 21
|
|
},
|
|
{
|
|
"epoch": 0.2816,
|
|
"grad_norm": 0.9353297733467271,
|
|
"learning_rate": 2.2564102564102566e-05,
|
|
"loss": 0.5431,
|
|
"step": 22
|
|
},
|
|
{
|
|
"epoch": 0.2944,
|
|
"grad_norm": 0.7191992110388274,
|
|
"learning_rate": 2.3589743589743593e-05,
|
|
"loss": 0.531,
|
|
"step": 23
|
|
},
|
|
{
|
|
"epoch": 0.3072,
|
|
"grad_norm": 0.7987803101137527,
|
|
"learning_rate": 2.461538461538462e-05,
|
|
"loss": 0.537,
|
|
"step": 24
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"grad_norm": 0.9000823684030751,
|
|
"learning_rate": 2.5641025641025646e-05,
|
|
"loss": 0.5243,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.3328,
|
|
"grad_norm": 0.639728529078365,
|
|
"learning_rate": 2.6666666666666667e-05,
|
|
"loss": 0.5256,
|
|
"step": 26
|
|
},
|
|
{
|
|
"epoch": 0.3456,
|
|
"grad_norm": 0.7310823848630157,
|
|
"learning_rate": 2.7692307692307694e-05,
|
|
"loss": 0.527,
|
|
"step": 27
|
|
},
|
|
{
|
|
"epoch": 0.3584,
|
|
"grad_norm": 0.7617952383130974,
|
|
"learning_rate": 2.871794871794872e-05,
|
|
"loss": 0.5243,
|
|
"step": 28
|
|
},
|
|
{
|
|
"epoch": 0.3712,
|
|
"grad_norm": 0.6087182615528468,
|
|
"learning_rate": 2.9743589743589747e-05,
|
|
"loss": 0.5213,
|
|
"step": 29
|
|
},
|
|
{
|
|
"epoch": 0.384,
|
|
"grad_norm": 0.6138496664446432,
|
|
"learning_rate": 3.0769230769230774e-05,
|
|
"loss": 0.5113,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.3968,
|
|
"grad_norm": 0.5842736668905947,
|
|
"learning_rate": 3.1794871794871795e-05,
|
|
"loss": 0.5088,
|
|
"step": 31
|
|
},
|
|
{
|
|
"epoch": 0.4096,
|
|
"grad_norm": 0.6842984878708783,
|
|
"learning_rate": 3.282051282051282e-05,
|
|
"loss": 0.5194,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.4224,
|
|
"grad_norm": 0.7772980067344607,
|
|
"learning_rate": 3.384615384615385e-05,
|
|
"loss": 0.5193,
|
|
"step": 33
|
|
},
|
|
{
|
|
"epoch": 0.4352,
|
|
"grad_norm": 0.6889382493737749,
|
|
"learning_rate": 3.4871794871794875e-05,
|
|
"loss": 0.5039,
|
|
"step": 34
|
|
},
|
|
{
|
|
"epoch": 0.448,
|
|
"grad_norm": 5.907781646563153,
|
|
"learning_rate": 3.58974358974359e-05,
|
|
"loss": 0.5234,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.4608,
|
|
"grad_norm": 1.833998012641166,
|
|
"learning_rate": 3.692307692307693e-05,
|
|
"loss": 0.5138,
|
|
"step": 36
|
|
},
|
|
{
|
|
"epoch": 0.4736,
|
|
"grad_norm": 1.358088445040064,
|
|
"learning_rate": 3.794871794871795e-05,
|
|
"loss": 0.5164,
|
|
"step": 37
|
|
},
|
|
{
|
|
"epoch": 0.4864,
|
|
"grad_norm": 0.9665299887078758,
|
|
"learning_rate": 3.8974358974358976e-05,
|
|
"loss": 0.5086,
|
|
"step": 38
|
|
},
|
|
{
|
|
"epoch": 0.4992,
|
|
"grad_norm": 1.8462558706888752,
|
|
"learning_rate": 4e-05,
|
|
"loss": 0.5108,
|
|
"step": 39
|
|
},
|
|
{
|
|
"epoch": 0.512,
|
|
"grad_norm": 1.0717311224964012,
|
|
"learning_rate": 3.9999198907597046e-05,
|
|
"loss": 0.4799,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.5248,
|
|
"grad_norm": 2.1436041861152764,
|
|
"learning_rate": 3.9996795694563096e-05,
|
|
"loss": 0.5216,
|
|
"step": 41
|
|
},
|
|
{
|
|
"epoch": 0.5376,
|
|
"grad_norm": 1.8190734353201568,
|
|
"learning_rate": 3.999279055341771e-05,
|
|
"loss": 0.4935,
|
|
"step": 42
|
|
},
|
|
{
|
|
"epoch": 0.5504,
|
|
"grad_norm": 1.3379628057404362,
|
|
"learning_rate": 3.998718380500971e-05,
|
|
"loss": 0.5215,
|
|
"step": 43
|
|
},
|
|
{
|
|
"epoch": 0.5632,
|
|
"grad_norm": 1.7052354362588542,
|
|
"learning_rate": 3.997997589849145e-05,
|
|
"loss": 0.5134,
|
|
"step": 44
|
|
},
|
|
{
|
|
"epoch": 0.576,
|
|
"grad_norm": 1.0266630844521263,
|
|
"learning_rate": 3.9971167411282835e-05,
|
|
"loss": 0.4923,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.5888,
|
|
"grad_norm": 1.3058420762275704,
|
|
"learning_rate": 3.99607590490251e-05,
|
|
"loss": 0.4989,
|
|
"step": 46
|
|
},
|
|
{
|
|
"epoch": 0.6016,
|
|
"grad_norm": 1.0731875195296174,
|
|
"learning_rate": 3.9948751645524235e-05,
|
|
"loss": 0.4942,
|
|
"step": 47
|
|
},
|
|
{
|
|
"epoch": 0.6144,
|
|
"grad_norm": 0.8898272327272633,
|
|
"learning_rate": 3.9935146162684206e-05,
|
|
"loss": 0.4992,
|
|
"step": 48
|
|
},
|
|
{
|
|
"epoch": 0.6272,
|
|
"grad_norm": 1.1040031356310187,
|
|
"learning_rate": 3.9919943690429906e-05,
|
|
"loss": 0.5025,
|
|
"step": 49
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"grad_norm": 0.8174445533574496,
|
|
"learning_rate": 3.9903145446619837e-05,
|
|
"loss": 0.4828,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.6528,
|
|
"grad_norm": 0.8673864554018242,
|
|
"learning_rate": 3.9884752776948564e-05,
|
|
"loss": 0.4791,
|
|
"step": 51
|
|
},
|
|
{
|
|
"epoch": 0.6656,
|
|
"grad_norm": 0.796148569165424,
|
|
"learning_rate": 3.9864767154838864e-05,
|
|
"loss": 0.4788,
|
|
"step": 52
|
|
},
|
|
{
|
|
"epoch": 0.6784,
|
|
"grad_norm": 0.6568989662572859,
|
|
"learning_rate": 3.9843190181323744e-05,
|
|
"loss": 0.4821,
|
|
"step": 53
|
|
},
|
|
{
|
|
"epoch": 0.6912,
|
|
"grad_norm": 0.8104547268528312,
|
|
"learning_rate": 3.982002358491817e-05,
|
|
"loss": 0.4838,
|
|
"step": 54
|
|
},
|
|
{
|
|
"epoch": 0.704,
|
|
"grad_norm": 0.6205373039374902,
|
|
"learning_rate": 3.979526922148058e-05,
|
|
"loss": 0.4977,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.7168,
|
|
"grad_norm": 0.7152522003873999,
|
|
"learning_rate": 3.9768929074064206e-05,
|
|
"loss": 0.476,
|
|
"step": 56
|
|
},
|
|
{
|
|
"epoch": 0.7296,
|
|
"grad_norm": 0.5502148010990727,
|
|
"learning_rate": 3.9741005252758255e-05,
|
|
"loss": 0.4851,
|
|
"step": 57
|
|
},
|
|
{
|
|
"epoch": 0.7424,
|
|
"grad_norm": 0.653670757127795,
|
|
"learning_rate": 3.971149999451886e-05,
|
|
"loss": 0.4758,
|
|
"step": 58
|
|
},
|
|
{
|
|
"epoch": 0.7552,
|
|
"grad_norm": 0.6467489669688496,
|
|
"learning_rate": 3.9680415662989806e-05,
|
|
"loss": 0.4708,
|
|
"step": 59
|
|
},
|
|
{
|
|
"epoch": 0.768,
|
|
"grad_norm": 0.7583281936365043,
|
|
"learning_rate": 3.9647754748313294e-05,
|
|
"loss": 0.4745,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.7808,
|
|
"grad_norm": 0.638240853677551,
|
|
"learning_rate": 3.96135198669304e-05,
|
|
"loss": 0.4855,
|
|
"step": 61
|
|
},
|
|
{
|
|
"epoch": 0.7936,
|
|
"grad_norm": 0.5490847741875153,
|
|
"learning_rate": 3.957771376137144e-05,
|
|
"loss": 0.4616,
|
|
"step": 62
|
|
},
|
|
{
|
|
"epoch": 0.8064,
|
|
"grad_norm": 0.5887261961332173,
|
|
"learning_rate": 3.954033930003634e-05,
|
|
"loss": 0.4751,
|
|
"step": 63
|
|
},
|
|
{
|
|
"epoch": 0.8192,
|
|
"grad_norm": 0.5124953087552118,
|
|
"learning_rate": 3.9501399476964806e-05,
|
|
"loss": 0.4767,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 0.832,
|
|
"grad_norm": 0.59821449783325,
|
|
"learning_rate": 3.946089741159648e-05,
|
|
"loss": 0.474,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.8448,
|
|
"grad_norm": 0.6120167845677714,
|
|
"learning_rate": 3.9418836348521045e-05,
|
|
"loss": 0.4749,
|
|
"step": 66
|
|
},
|
|
{
|
|
"epoch": 0.8576,
|
|
"grad_norm": 0.5235563985592369,
|
|
"learning_rate": 3.937521965721831e-05,
|
|
"loss": 0.4748,
|
|
"step": 67
|
|
},
|
|
{
|
|
"epoch": 0.8704,
|
|
"grad_norm": 0.5350857351039692,
|
|
"learning_rate": 3.933005083178828e-05,
|
|
"loss": 0.4673,
|
|
"step": 68
|
|
},
|
|
{
|
|
"epoch": 0.8832,
|
|
"grad_norm": 0.6142423277054795,
|
|
"learning_rate": 3.928333349067125e-05,
|
|
"loss": 0.4767,
|
|
"step": 69
|
|
},
|
|
{
|
|
"epoch": 0.896,
|
|
"grad_norm": 0.590492201343943,
|
|
"learning_rate": 3.923507137635792e-05,
|
|
"loss": 0.4679,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.9088,
|
|
"grad_norm": 0.47391904061487716,
|
|
"learning_rate": 3.9185268355089606e-05,
|
|
"loss": 0.465,
|
|
"step": 71
|
|
},
|
|
{
|
|
"epoch": 0.9216,
|
|
"grad_norm": 0.5617478093615034,
|
|
"learning_rate": 3.913392841654851e-05,
|
|
"loss": 0.4735,
|
|
"step": 72
|
|
},
|
|
{
|
|
"epoch": 0.9344,
|
|
"grad_norm": 0.47206063489404443,
|
|
"learning_rate": 3.9081055673538093e-05,
|
|
"loss": 0.4697,
|
|
"step": 73
|
|
},
|
|
{
|
|
"epoch": 0.9472,
|
|
"grad_norm": 0.5078461387459143,
|
|
"learning_rate": 3.902665436165364e-05,
|
|
"loss": 0.4752,
|
|
"step": 74
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"grad_norm": 0.5108689431705631,
|
|
"learning_rate": 3.897072883894291e-05,
|
|
"loss": 0.4484,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.9728,
|
|
"grad_norm": 0.5288717566865853,
|
|
"learning_rate": 3.8913283585557054e-05,
|
|
"loss": 0.4584,
|
|
"step": 76
|
|
},
|
|
{
|
|
"epoch": 0.9856,
|
|
"grad_norm": 0.5106573654152327,
|
|
"learning_rate": 3.885432320339167e-05,
|
|
"loss": 0.4695,
|
|
"step": 77
|
|
},
|
|
{
|
|
"epoch": 0.9984,
|
|
"grad_norm": 0.4978461545439326,
|
|
"learning_rate": 3.879385241571817e-05,
|
|
"loss": 0.4659,
|
|
"step": 78
|
|
},
|
|
{
|
|
"epoch": 1.0112,
|
|
"grad_norm": 0.5374431536494356,
|
|
"learning_rate": 3.873187606680543e-05,
|
|
"loss": 0.4383,
|
|
"step": 79
|
|
},
|
|
{
|
|
"epoch": 1.024,
|
|
"grad_norm": 0.466905661747252,
|
|
"learning_rate": 3.866839912153168e-05,
|
|
"loss": 0.4235,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 1.0368,
|
|
"grad_norm": 0.5524070933573034,
|
|
"learning_rate": 3.860342666498677e-05,
|
|
"loss": 0.426,
|
|
"step": 81
|
|
},
|
|
{
|
|
"epoch": 1.0496,
|
|
"grad_norm": 0.5552329501082961,
|
|
"learning_rate": 3.853696390206484e-05,
|
|
"loss": 0.4293,
|
|
"step": 82
|
|
},
|
|
{
|
|
"epoch": 1.0624,
|
|
"grad_norm": 0.6365882933312345,
|
|
"learning_rate": 3.846901615704734e-05,
|
|
"loss": 0.4277,
|
|
"step": 83
|
|
},
|
|
{
|
|
"epoch": 1.0752,
|
|
"grad_norm": 0.6792621794760965,
|
|
"learning_rate": 3.839958887317649e-05,
|
|
"loss": 0.4136,
|
|
"step": 84
|
|
},
|
|
{
|
|
"epoch": 1.088,
|
|
"grad_norm": 0.7086948729956156,
|
|
"learning_rate": 3.832868761221926e-05,
|
|
"loss": 0.4203,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 1.1008,
|
|
"grad_norm": 0.5914674949977854,
|
|
"learning_rate": 3.825631805402182e-05,
|
|
"loss": 0.4133,
|
|
"step": 86
|
|
},
|
|
{
|
|
"epoch": 1.1136,
|
|
"grad_norm": 0.6457869945763296,
|
|
"learning_rate": 3.818248599605448e-05,
|
|
"loss": 0.4189,
|
|
"step": 87
|
|
},
|
|
{
|
|
"epoch": 1.1264,
|
|
"grad_norm": 0.7794506477481747,
|
|
"learning_rate": 3.810719735294731e-05,
|
|
"loss": 0.4319,
|
|
"step": 88
|
|
},
|
|
{
|
|
"epoch": 1.1392,
|
|
"grad_norm": 0.6157599617077694,
|
|
"learning_rate": 3.8030458156016326e-05,
|
|
"loss": 0.4138,
|
|
"step": 89
|
|
},
|
|
{
|
|
"epoch": 1.152,
|
|
"grad_norm": 0.5668112614414113,
|
|
"learning_rate": 3.795227455278029e-05,
|
|
"loss": 0.4087,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 1.1648,
|
|
"grad_norm": 0.7534233323894747,
|
|
"learning_rate": 3.787265280646825e-05,
|
|
"loss": 0.4158,
|
|
"step": 91
|
|
},
|
|
{
|
|
"epoch": 1.1776,
|
|
"grad_norm": 0.8557762634468793,
|
|
"learning_rate": 3.7791599295517825e-05,
|
|
"loss": 0.4163,
|
|
"step": 92
|
|
},
|
|
{
|
|
"epoch": 1.1904,
|
|
"grad_norm": 0.5514378853304944,
|
|
"learning_rate": 3.7709120513064196e-05,
|
|
"loss": 0.4166,
|
|
"step": 93
|
|
},
|
|
{
|
|
"epoch": 1.2032,
|
|
"grad_norm": 0.6176747491282486,
|
|
"learning_rate": 3.762522306641998e-05,
|
|
"loss": 0.4103,
|
|
"step": 94
|
|
},
|
|
{
|
|
"epoch": 1.216,
|
|
"grad_norm": 0.7020404535006932,
|
|
"learning_rate": 3.7539913676545874e-05,
|
|
"loss": 0.3998,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 1.2288000000000001,
|
|
"grad_norm": 0.36522429007528184,
|
|
"learning_rate": 3.745319917751229e-05,
|
|
"loss": 0.4097,
|
|
"step": 96
|
|
},
|
|
{
|
|
"epoch": 1.2416,
|
|
"grad_norm": 0.5900634589160236,
|
|
"learning_rate": 3.736508651595188e-05,
|
|
"loss": 0.4228,
|
|
"step": 97
|
|
},
|
|
{
|
|
"epoch": 1.2544,
|
|
"grad_norm": 0.4613833081509036,
|
|
"learning_rate": 3.727558275050301e-05,
|
|
"loss": 0.4124,
|
|
"step": 98
|
|
},
|
|
{
|
|
"epoch": 1.2671999999999999,
|
|
"grad_norm": 0.4846729996977901,
|
|
"learning_rate": 3.718469505124434e-05,
|
|
"loss": 0.404,
|
|
"step": 99
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"grad_norm": 0.4683152886505894,
|
|
"learning_rate": 3.709243069912041e-05,
|
|
"loss": 0.4187,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 1.2928,
|
|
"grad_norm": 0.3717197380076313,
|
|
"learning_rate": 3.699879708535838e-05,
|
|
"loss": 0.4139,
|
|
"step": 101
|
|
},
|
|
{
|
|
"epoch": 1.3056,
|
|
"grad_norm": 0.4302258236247511,
|
|
"learning_rate": 3.69038017108759e-05,
|
|
"loss": 0.4159,
|
|
"step": 102
|
|
},
|
|
{
|
|
"epoch": 1.3184,
|
|
"grad_norm": 0.3982962645464779,
|
|
"learning_rate": 3.680745218568026e-05,
|
|
"loss": 0.4181,
|
|
"step": 103
|
|
},
|
|
{
|
|
"epoch": 1.3312,
|
|
"grad_norm": 0.40200214235666887,
|
|
"learning_rate": 3.6709756228258735e-05,
|
|
"loss": 0.4046,
|
|
"step": 104
|
|
},
|
|
{
|
|
"epoch": 1.3439999999999999,
|
|
"grad_norm": 0.44402430474888865,
|
|
"learning_rate": 3.6610721664960236e-05,
|
|
"loss": 0.4263,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 1.3568,
|
|
"grad_norm": 0.38820837732347074,
|
|
"learning_rate": 3.65103564293684e-05,
|
|
"loss": 0.417,
|
|
"step": 106
|
|
},
|
|
{
|
|
"epoch": 1.3696,
|
|
"grad_norm": 0.5213458360363288,
|
|
"learning_rate": 3.640866856166601e-05,
|
|
"loss": 0.426,
|
|
"step": 107
|
|
},
|
|
{
|
|
"epoch": 1.3824,
|
|
"grad_norm": 0.4308788707291437,
|
|
"learning_rate": 3.6305666207990886e-05,
|
|
"loss": 0.4097,
|
|
"step": 108
|
|
},
|
|
{
|
|
"epoch": 1.3952,
|
|
"grad_norm": 0.4644795634740679,
|
|
"learning_rate": 3.6201357619783336e-05,
|
|
"loss": 0.4068,
|
|
"step": 109
|
|
},
|
|
{
|
|
"epoch": 1.408,
|
|
"grad_norm": 0.44267654288374053,
|
|
"learning_rate": 3.609575115312511e-05,
|
|
"loss": 0.415,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 1.4208,
|
|
"grad_norm": 0.4714173326270217,
|
|
"learning_rate": 3.598885526807003e-05,
|
|
"loss": 0.4101,
|
|
"step": 111
|
|
},
|
|
{
|
|
"epoch": 1.4336,
|
|
"grad_norm": 0.4805820675277153,
|
|
"learning_rate": 3.5880678527966224e-05,
|
|
"loss": 0.4213,
|
|
"step": 112
|
|
},
|
|
{
|
|
"epoch": 1.4464000000000001,
|
|
"grad_norm": 0.39641366601935735,
|
|
"learning_rate": 3.577122959877017e-05,
|
|
"loss": 0.3989,
|
|
"step": 113
|
|
},
|
|
{
|
|
"epoch": 1.4592,
|
|
"grad_norm": 0.3759641748474472,
|
|
"learning_rate": 3.566051724835245e-05,
|
|
"loss": 0.416,
|
|
"step": 114
|
|
},
|
|
{
|
|
"epoch": 1.472,
|
|
"grad_norm": 0.3694282519820903,
|
|
"learning_rate": 3.554855034579532e-05,
|
|
"loss": 0.4225,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 1.4848,
|
|
"grad_norm": 0.4024973231579404,
|
|
"learning_rate": 3.5435337860682304e-05,
|
|
"loss": 0.4153,
|
|
"step": 116
|
|
},
|
|
{
|
|
"epoch": 1.4976,
|
|
"grad_norm": 0.37364740883863296,
|
|
"learning_rate": 3.532088886237956e-05,
|
|
"loss": 0.4108,
|
|
"step": 117
|
|
},
|
|
{
|
|
"epoch": 1.5104,
|
|
"grad_norm": 0.36906546048784605,
|
|
"learning_rate": 3.520521251930941e-05,
|
|
"loss": 0.4124,
|
|
"step": 118
|
|
},
|
|
{
|
|
"epoch": 1.5232,
|
|
"grad_norm": 0.3748555038416534,
|
|
"learning_rate": 3.5088318098215805e-05,
|
|
"loss": 0.4081,
|
|
"step": 119
|
|
},
|
|
{
|
|
"epoch": 1.536,
|
|
"grad_norm": 0.3582549157315094,
|
|
"learning_rate": 3.497021496342203e-05,
|
|
"loss": 0.4018,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 1.5488,
|
|
"grad_norm": 0.3887317377098141,
|
|
"learning_rate": 3.485091257608047e-05,
|
|
"loss": 0.3943,
|
|
"step": 121
|
|
},
|
|
{
|
|
"epoch": 1.5615999999999999,
|
|
"grad_norm": 0.4585279897491825,
|
|
"learning_rate": 3.473042049341474e-05,
|
|
"loss": 0.4001,
|
|
"step": 122
|
|
},
|
|
{
|
|
"epoch": 1.5744,
|
|
"grad_norm": 0.37694477337454396,
|
|
"learning_rate": 3.4608748367954064e-05,
|
|
"loss": 0.4002,
|
|
"step": 123
|
|
},
|
|
{
|
|
"epoch": 1.5872000000000002,
|
|
"grad_norm": 0.39032755502498523,
|
|
"learning_rate": 3.4485905946759965e-05,
|
|
"loss": 0.4078,
|
|
"step": 124
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"grad_norm": 0.4322350994624392,
|
|
"learning_rate": 3.4361903070645484e-05,
|
|
"loss": 0.4025,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 1.6128,
|
|
"grad_norm": 0.39632556623882664,
|
|
"learning_rate": 3.423674967338681e-05,
|
|
"loss": 0.4062,
|
|
"step": 126
|
|
},
|
|
{
|
|
"epoch": 1.6256,
|
|
"grad_norm": 0.41948444365675525,
|
|
"learning_rate": 3.411045578092754e-05,
|
|
"loss": 0.4177,
|
|
"step": 127
|
|
},
|
|
{
|
|
"epoch": 1.6383999999999999,
|
|
"grad_norm": 0.4799383655202944,
|
|
"learning_rate": 3.398303151057543e-05,
|
|
"loss": 0.4166,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 1.6512,
|
|
"grad_norm": 0.47130088781959734,
|
|
"learning_rate": 3.385448707019199e-05,
|
|
"loss": 0.4095,
|
|
"step": 129
|
|
},
|
|
{
|
|
"epoch": 1.6640000000000001,
|
|
"grad_norm": 0.38829379832198224,
|
|
"learning_rate": 3.372483275737468e-05,
|
|
"loss": 0.4088,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 1.6768,
|
|
"grad_norm": 0.48425044375595544,
|
|
"learning_rate": 3.359407895863199e-05,
|
|
"loss": 0.4203,
|
|
"step": 131
|
|
},
|
|
{
|
|
"epoch": 1.6896,
|
|
"grad_norm": 0.5022610534269328,
|
|
"learning_rate": 3.34622361485514e-05,
|
|
"loss": 0.4049,
|
|
"step": 132
|
|
},
|
|
{
|
|
"epoch": 1.7024,
|
|
"grad_norm": 0.43114127314574263,
|
|
"learning_rate": 3.332931488896029e-05,
|
|
"loss": 0.4093,
|
|
"step": 133
|
|
},
|
|
{
|
|
"epoch": 1.7151999999999998,
|
|
"grad_norm": 0.5129917681391781,
|
|
"learning_rate": 3.319532582807977e-05,
|
|
"loss": 0.4161,
|
|
"step": 134
|
|
},
|
|
{
|
|
"epoch": 1.728,
|
|
"grad_norm": 0.4853215798673184,
|
|
"learning_rate": 3.30602796996717e-05,
|
|
"loss": 0.4142,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 1.7408000000000001,
|
|
"grad_norm": 0.5395589871823466,
|
|
"learning_rate": 3.2924187322178865e-05,
|
|
"loss": 0.4162,
|
|
"step": 136
|
|
},
|
|
{
|
|
"epoch": 1.7536,
|
|
"grad_norm": 0.40618095402028487,
|
|
"learning_rate": 3.278705959785821e-05,
|
|
"loss": 0.4251,
|
|
"step": 137
|
|
},
|
|
{
|
|
"epoch": 1.7664,
|
|
"grad_norm": 0.4746670426025735,
|
|
"learning_rate": 3.2648907511907544e-05,
|
|
"loss": 0.4018,
|
|
"step": 138
|
|
},
|
|
{
|
|
"epoch": 1.7792,
|
|
"grad_norm": 0.38240123980896445,
|
|
"learning_rate": 3.250974213158555e-05,
|
|
"loss": 0.4096,
|
|
"step": 139
|
|
},
|
|
{
|
|
"epoch": 1.792,
|
|
"grad_norm": 0.4587704974203193,
|
|
"learning_rate": 3.23695746053251e-05,
|
|
"loss": 0.4142,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 1.8048,
|
|
"grad_norm": 0.4138827354677887,
|
|
"learning_rate": 3.222841616184025e-05,
|
|
"loss": 0.418,
|
|
"step": 141
|
|
},
|
|
{
|
|
"epoch": 1.8176,
|
|
"grad_norm": 0.45435877514938233,
|
|
"learning_rate": 3.208627810922665e-05,
|
|
"loss": 0.4028,
|
|
"step": 142
|
|
},
|
|
{
|
|
"epoch": 1.8304,
|
|
"grad_norm": 0.4647800854837698,
|
|
"learning_rate": 3.194317183405573e-05,
|
|
"loss": 0.4132,
|
|
"step": 143
|
|
},
|
|
{
|
|
"epoch": 1.8432,
|
|
"grad_norm": 0.36420373665662603,
|
|
"learning_rate": 3.1799108800462466e-05,
|
|
"loss": 0.4031,
|
|
"step": 144
|
|
},
|
|
{
|
|
"epoch": 1.8559999999999999,
|
|
"grad_norm": 0.4666313110580902,
|
|
"learning_rate": 3.1654100549227024e-05,
|
|
"loss": 0.4162,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 1.8688,
|
|
"grad_norm": 0.39777290497002804,
|
|
"learning_rate": 3.1508158696850275e-05,
|
|
"loss": 0.4015,
|
|
"step": 146
|
|
},
|
|
{
|
|
"epoch": 1.8816000000000002,
|
|
"grad_norm": 0.41964971235728266,
|
|
"learning_rate": 3.136129493462312e-05,
|
|
"loss": 0.4051,
|
|
"step": 147
|
|
},
|
|
{
|
|
"epoch": 1.8944,
|
|
"grad_norm": 0.42487362870525114,
|
|
"learning_rate": 3.121352102768998e-05,
|
|
"loss": 0.4121,
|
|
"step": 148
|
|
},
|
|
{
|
|
"epoch": 1.9072,
|
|
"grad_norm": 0.34253627904746686,
|
|
"learning_rate": 3.106484881410628e-05,
|
|
"loss": 0.4126,
|
|
"step": 149
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"grad_norm": 0.3822674020304617,
|
|
"learning_rate": 3.091529020389009e-05,
|
|
"loss": 0.4131,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 1.9327999999999999,
|
|
"grad_norm": 0.40435918024987844,
|
|
"learning_rate": 3.076485717806808e-05,
|
|
"loss": 0.4148,
|
|
"step": 151
|
|
},
|
|
{
|
|
"epoch": 1.9456,
|
|
"grad_norm": 0.3499098381687515,
|
|
"learning_rate": 3.061356178771564e-05,
|
|
"loss": 0.3936,
|
|
"step": 152
|
|
},
|
|
{
|
|
"epoch": 1.9584000000000001,
|
|
"grad_norm": 0.4076701390830936,
|
|
"learning_rate": 3.0461416152991555e-05,
|
|
"loss": 0.4075,
|
|
"step": 153
|
|
},
|
|
{
|
|
"epoch": 1.9712,
|
|
"grad_norm": 0.40087378483854863,
|
|
"learning_rate": 3.0308432462167045e-05,
|
|
"loss": 0.4057,
|
|
"step": 154
|
|
},
|
|
{
|
|
"epoch": 1.984,
|
|
"grad_norm": 0.41603083979984834,
|
|
"learning_rate": 3.015462297064936e-05,
|
|
"loss": 0.4026,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 1.9968,
|
|
"grad_norm": 0.3504122558670022,
|
|
"learning_rate": 3.0000000000000004e-05,
|
|
"loss": 0.3867,
|
|
"step": 156
|
|
},
|
|
{
|
|
"epoch": 2.0096,
|
|
"grad_norm": 0.5272796148664116,
|
|
"learning_rate": 2.98445759369477e-05,
|
|
"loss": 0.3827,
|
|
"step": 157
|
|
},
|
|
{
|
|
"epoch": 2.0224,
|
|
"grad_norm": 0.44531053711387253,
|
|
"learning_rate": 2.9688363232396056e-05,
|
|
"loss": 0.3576,
|
|
"step": 158
|
|
},
|
|
{
|
|
"epoch": 2.0352,
|
|
"grad_norm": 0.5959308354780575,
|
|
"learning_rate": 2.9531374400426158e-05,
|
|
"loss": 0.358,
|
|
"step": 159
|
|
},
|
|
{
|
|
"epoch": 2.048,
|
|
"grad_norm": 0.5693188685590685,
|
|
"learning_rate": 2.9373622017294075e-05,
|
|
"loss": 0.3447,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 2.0608,
|
|
"grad_norm": 0.5042613043316267,
|
|
"learning_rate": 2.9215118720423375e-05,
|
|
"loss": 0.3417,
|
|
"step": 161
|
|
},
|
|
{
|
|
"epoch": 2.0736,
|
|
"grad_norm": 0.553051231213321,
|
|
"learning_rate": 2.9055877207392752e-05,
|
|
"loss": 0.3474,
|
|
"step": 162
|
|
},
|
|
{
|
|
"epoch": 2.0864,
|
|
"grad_norm": 0.4557734083518579,
|
|
"learning_rate": 2.8895910234918828e-05,
|
|
"loss": 0.3546,
|
|
"step": 163
|
|
},
|
|
{
|
|
"epoch": 2.0992,
|
|
"grad_norm": 0.5416631636159364,
|
|
"learning_rate": 2.873523061783426e-05,
|
|
"loss": 0.3483,
|
|
"step": 164
|
|
},
|
|
{
|
|
"epoch": 2.112,
|
|
"grad_norm": 0.5389527047734003,
|
|
"learning_rate": 2.8573851228061084e-05,
|
|
"loss": 0.341,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 2.1248,
|
|
"grad_norm": 0.4105182765148673,
|
|
"learning_rate": 2.8411784993579633e-05,
|
|
"loss": 0.3389,
|
|
"step": 166
|
|
},
|
|
{
|
|
"epoch": 2.1376,
|
|
"grad_norm": 0.56488782807409,
|
|
"learning_rate": 2.8249044897392814e-05,
|
|
"loss": 0.3473,
|
|
"step": 167
|
|
},
|
|
{
|
|
"epoch": 2.1504,
|
|
"grad_norm": 0.3973328446053959,
|
|
"learning_rate": 2.80856439764861e-05,
|
|
"loss": 0.3393,
|
|
"step": 168
|
|
},
|
|
{
|
|
"epoch": 2.1632,
|
|
"grad_norm": 0.48655927946324257,
|
|
"learning_rate": 2.792159532078314e-05,
|
|
"loss": 0.3409,
|
|
"step": 169
|
|
},
|
|
{
|
|
"epoch": 2.176,
|
|
"grad_norm": 0.3562939309608415,
|
|
"learning_rate": 2.77569120720971e-05,
|
|
"loss": 0.3339,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 2.1888,
|
|
"grad_norm": 0.3865349086129511,
|
|
"learning_rate": 2.7591607423077932e-05,
|
|
"loss": 0.3441,
|
|
"step": 171
|
|
},
|
|
{
|
|
"epoch": 2.2016,
|
|
"grad_norm": 0.38384389039352235,
|
|
"learning_rate": 2.7425694616155474e-05,
|
|
"loss": 0.3413,
|
|
"step": 172
|
|
},
|
|
{
|
|
"epoch": 2.2144,
|
|
"grad_norm": 0.32530141741650587,
|
|
"learning_rate": 2.7259186942478656e-05,
|
|
"loss": 0.3455,
|
|
"step": 173
|
|
},
|
|
{
|
|
"epoch": 2.2272,
|
|
"grad_norm": 0.3686673261248661,
|
|
"learning_rate": 2.7092097740850712e-05,
|
|
"loss": 0.3481,
|
|
"step": 174
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"grad_norm": 0.3143044879080098,
|
|
"learning_rate": 2.692444039666066e-05,
|
|
"loss": 0.3337,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 2.2528,
|
|
"grad_norm": 0.31619399123559355,
|
|
"learning_rate": 2.6756228340810946e-05,
|
|
"loss": 0.3423,
|
|
"step": 176
|
|
},
|
|
{
|
|
"epoch": 2.2656,
|
|
"grad_norm": 0.3373730218823642,
|
|
"learning_rate": 2.6587475048641596e-05,
|
|
"loss": 0.3466,
|
|
"step": 177
|
|
},
|
|
{
|
|
"epoch": 2.2784,
|
|
"grad_norm": 0.3409092059873659,
|
|
"learning_rate": 2.6418194038850634e-05,
|
|
"loss": 0.3401,
|
|
"step": 178
|
|
},
|
|
{
|
|
"epoch": 2.2912,
|
|
"grad_norm": 0.3262535081367611,
|
|
"learning_rate": 2.624839887241115e-05,
|
|
"loss": 0.3355,
|
|
"step": 179
|
|
},
|
|
{
|
|
"epoch": 2.304,
|
|
"grad_norm": 0.3598189947105674,
|
|
"learning_rate": 2.607810315148494e-05,
|
|
"loss": 0.3384,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 2.3168,
|
|
"grad_norm": 0.3082225835028636,
|
|
"learning_rate": 2.5907320518332827e-05,
|
|
"loss": 0.3434,
|
|
"step": 181
|
|
},
|
|
{
|
|
"epoch": 2.3296,
|
|
"grad_norm": 0.37794853145832014,
|
|
"learning_rate": 2.5736064654221808e-05,
|
|
"loss": 0.347,
|
|
"step": 182
|
|
},
|
|
{
|
|
"epoch": 2.3424,
|
|
"grad_norm": 0.3441245055128099,
|
|
"learning_rate": 2.5564349278329056e-05,
|
|
"loss": 0.3434,
|
|
"step": 183
|
|
},
|
|
{
|
|
"epoch": 2.3552,
|
|
"grad_norm": 0.34812376784545485,
|
|
"learning_rate": 2.539218814664288e-05,
|
|
"loss": 0.336,
|
|
"step": 184
|
|
},
|
|
{
|
|
"epoch": 2.368,
|
|
"grad_norm": 0.3611431416358283,
|
|
"learning_rate": 2.521959505086075e-05,
|
|
"loss": 0.3431,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 2.3808,
|
|
"grad_norm": 0.4222133844001757,
|
|
"learning_rate": 2.5046583817284437e-05,
|
|
"loss": 0.3373,
|
|
"step": 186
|
|
},
|
|
{
|
|
"epoch": 2.3936,
|
|
"grad_norm": 0.31914592506777995,
|
|
"learning_rate": 2.487316830571244e-05,
|
|
"loss": 0.3389,
|
|
"step": 187
|
|
},
|
|
{
|
|
"epoch": 2.4064,
|
|
"grad_norm": 0.31451553979055025,
|
|
"learning_rate": 2.4699362408329646e-05,
|
|
"loss": 0.3303,
|
|
"step": 188
|
|
},
|
|
{
|
|
"epoch": 2.4192,
|
|
"grad_norm": 0.37003768460022973,
|
|
"learning_rate": 2.4525180048594452e-05,
|
|
"loss": 0.3502,
|
|
"step": 189
|
|
},
|
|
{
|
|
"epoch": 2.432,
|
|
"grad_norm": 0.3239873799914562,
|
|
"learning_rate": 2.435063518012335e-05,
|
|
"loss": 0.342,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 2.4448,
|
|
"grad_norm": 0.31332443912294144,
|
|
"learning_rate": 2.4175741785573177e-05,
|
|
"loss": 0.3486,
|
|
"step": 191
|
|
},
|
|
{
|
|
"epoch": 2.4576000000000002,
|
|
"grad_norm": 0.3394957631642956,
|
|
"learning_rate": 2.4000513875520892e-05,
|
|
"loss": 0.339,
|
|
"step": 192
|
|
},
|
|
{
|
|
"epoch": 2.4704,
|
|
"grad_norm": 0.30179131299529555,
|
|
"learning_rate": 2.3824965487341247e-05,
|
|
"loss": 0.3346,
|
|
"step": 193
|
|
},
|
|
{
|
|
"epoch": 2.4832,
|
|
"grad_norm": 0.311970573179333,
|
|
"learning_rate": 2.3649110684082258e-05,
|
|
"loss": 0.3413,
|
|
"step": 194
|
|
},
|
|
{
|
|
"epoch": 2.496,
|
|
"grad_norm": 0.30637190878979437,
|
|
"learning_rate": 2.3472963553338614e-05,
|
|
"loss": 0.3415,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 2.5088,
|
|
"grad_norm": 0.3262722361133359,
|
|
"learning_rate": 2.3296538206123134e-05,
|
|
"loss": 0.3424,
|
|
"step": 196
|
|
},
|
|
{
|
|
"epoch": 2.5216,
|
|
"grad_norm": 0.30807470653995994,
|
|
"learning_rate": 2.311984877573636e-05,
|
|
"loss": 0.3308,
|
|
"step": 197
|
|
},
|
|
{
|
|
"epoch": 2.5343999999999998,
|
|
"grad_norm": 0.334927287671144,
|
|
"learning_rate": 2.2942909416634326e-05,
|
|
"loss": 0.3457,
|
|
"step": 198
|
|
},
|
|
{
|
|
"epoch": 2.5472,
|
|
"grad_norm": 0.28967982177015333,
|
|
"learning_rate": 2.2765734303294666e-05,
|
|
"loss": 0.3428,
|
|
"step": 199
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"grad_norm": 0.36898835911684036,
|
|
"learning_rate": 2.2588337629081107e-05,
|
|
"loss": 0.3406,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 2.5728,
|
|
"grad_norm": 0.28145488084817705,
|
|
"learning_rate": 2.2410733605106462e-05,
|
|
"loss": 0.3452,
|
|
"step": 201
|
|
},
|
|
{
|
|
"epoch": 2.5856,
|
|
"grad_norm": 0.3941864938083129,
|
|
"learning_rate": 2.2232936459094158e-05,
|
|
"loss": 0.3399,
|
|
"step": 202
|
|
},
|
|
{
|
|
"epoch": 2.5984,
|
|
"grad_norm": 0.31259268252215316,
|
|
"learning_rate": 2.205496043423849e-05,
|
|
"loss": 0.3355,
|
|
"step": 203
|
|
},
|
|
{
|
|
"epoch": 2.6112,
|
|
"grad_norm": 0.37615122244959603,
|
|
"learning_rate": 2.1876819788063586e-05,
|
|
"loss": 0.3394,
|
|
"step": 204
|
|
},
|
|
{
|
|
"epoch": 2.624,
|
|
"grad_norm": 0.2940297985807267,
|
|
"learning_rate": 2.16985287912813e-05,
|
|
"loss": 0.3458,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 2.6368,
|
|
"grad_norm": 0.4070811437427959,
|
|
"learning_rate": 2.1520101726647922e-05,
|
|
"loss": 0.3427,
|
|
"step": 206
|
|
},
|
|
{
|
|
"epoch": 2.6496,
|
|
"grad_norm": 0.32272328892363317,
|
|
"learning_rate": 2.1341552887820048e-05,
|
|
"loss": 0.341,
|
|
"step": 207
|
|
},
|
|
{
|
|
"epoch": 2.6624,
|
|
"grad_norm": 0.4474845128230661,
|
|
"learning_rate": 2.1162896578209517e-05,
|
|
"loss": 0.3451,
|
|
"step": 208
|
|
},
|
|
{
|
|
"epoch": 2.6752000000000002,
|
|
"grad_norm": 0.3499793256711866,
|
|
"learning_rate": 2.0984147109837564e-05,
|
|
"loss": 0.341,
|
|
"step": 209
|
|
},
|
|
{
|
|
"epoch": 2.6879999999999997,
|
|
"grad_norm": 0.39384147774941897,
|
|
"learning_rate": 2.0805318802188307e-05,
|
|
"loss": 0.3499,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 2.7008,
|
|
"grad_norm": 0.35193869889879736,
|
|
"learning_rate": 2.0626425981061608e-05,
|
|
"loss": 0.337,
|
|
"step": 211
|
|
},
|
|
{
|
|
"epoch": 2.7136,
|
|
"grad_norm": 0.3726248118838429,
|
|
"learning_rate": 2.0447482977425465e-05,
|
|
"loss": 0.3364,
|
|
"step": 212
|
|
},
|
|
{
|
|
"epoch": 2.7264,
|
|
"grad_norm": 0.2919938332617151,
|
|
"learning_rate": 2.0268504126267952e-05,
|
|
"loss": 0.339,
|
|
"step": 213
|
|
},
|
|
{
|
|
"epoch": 2.7392,
|
|
"grad_norm": 0.3339209869689389,
|
|
"learning_rate": 2.008950376544887e-05,
|
|
"loss": 0.3425,
|
|
"step": 214
|
|
},
|
|
{
|
|
"epoch": 2.752,
|
|
"grad_norm": 0.29640305737764455,
|
|
"learning_rate": 1.9910496234551132e-05,
|
|
"loss": 0.3402,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 2.7648,
|
|
"grad_norm": 0.30254409880624555,
|
|
"learning_rate": 1.9731495873732055e-05,
|
|
"loss": 0.3531,
|
|
"step": 216
|
|
},
|
|
{
|
|
"epoch": 2.7776,
|
|
"grad_norm": 0.30911087112319563,
|
|
"learning_rate": 1.9552517022574542e-05,
|
|
"loss": 0.343,
|
|
"step": 217
|
|
},
|
|
{
|
|
"epoch": 2.7904,
|
|
"grad_norm": 0.2563063478891904,
|
|
"learning_rate": 1.93735740189384e-05,
|
|
"loss": 0.3449,
|
|
"step": 218
|
|
},
|
|
{
|
|
"epoch": 2.8032,
|
|
"grad_norm": 0.2705467137959849,
|
|
"learning_rate": 1.9194681197811703e-05,
|
|
"loss": 0.3569,
|
|
"step": 219
|
|
},
|
|
{
|
|
"epoch": 2.816,
|
|
"grad_norm": 0.2682899014111958,
|
|
"learning_rate": 1.901585289016244e-05,
|
|
"loss": 0.3338,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 2.8288,
|
|
"grad_norm": 0.27996514808139306,
|
|
"learning_rate": 1.8837103421790486e-05,
|
|
"loss": 0.3316,
|
|
"step": 221
|
|
},
|
|
{
|
|
"epoch": 2.8416,
|
|
"grad_norm": 0.25978948174639804,
|
|
"learning_rate": 1.8658447112179952e-05,
|
|
"loss": 0.3326,
|
|
"step": 222
|
|
},
|
|
{
|
|
"epoch": 2.8544,
|
|
"grad_norm": 0.2564459029629428,
|
|
"learning_rate": 1.8479898273352084e-05,
|
|
"loss": 0.3269,
|
|
"step": 223
|
|
},
|
|
{
|
|
"epoch": 2.8672,
|
|
"grad_norm": 0.2854711673452058,
|
|
"learning_rate": 1.83014712087187e-05,
|
|
"loss": 0.3437,
|
|
"step": 224
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"grad_norm": 0.27320772787317565,
|
|
"learning_rate": 1.8123180211936417e-05,
|
|
"loss": 0.3317,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 2.8928000000000003,
|
|
"grad_norm": 0.2966590212320321,
|
|
"learning_rate": 1.794503956576152e-05,
|
|
"loss": 0.3391,
|
|
"step": 226
|
|
},
|
|
{
|
|
"epoch": 2.9055999999999997,
|
|
"grad_norm": 0.31906514403005093,
|
|
"learning_rate": 1.776706354090585e-05,
|
|
"loss": 0.349,
|
|
"step": 227
|
|
},
|
|
{
|
|
"epoch": 2.9184,
|
|
"grad_norm": 0.2749646812104898,
|
|
"learning_rate": 1.758926639489354e-05,
|
|
"loss": 0.3435,
|
|
"step": 228
|
|
},
|
|
{
|
|
"epoch": 2.9312,
|
|
"grad_norm": 0.34925561830239105,
|
|
"learning_rate": 1.7411662370918893e-05,
|
|
"loss": 0.3367,
|
|
"step": 229
|
|
},
|
|
{
|
|
"epoch": 2.944,
|
|
"grad_norm": 0.3258150309203499,
|
|
"learning_rate": 1.7234265696705344e-05,
|
|
"loss": 0.3398,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 2.9568,
|
|
"grad_norm": 0.31046565035409324,
|
|
"learning_rate": 1.7057090583365678e-05,
|
|
"loss": 0.3401,
|
|
"step": 231
|
|
},
|
|
{
|
|
"epoch": 2.9696,
|
|
"grad_norm": 0.27678310267277134,
|
|
"learning_rate": 1.6880151224263646e-05,
|
|
"loss": 0.3468,
|
|
"step": 232
|
|
},
|
|
{
|
|
"epoch": 2.9824,
|
|
"grad_norm": 0.26358082043532155,
|
|
"learning_rate": 1.6703461793876876e-05,
|
|
"loss": 0.3455,
|
|
"step": 233
|
|
},
|
|
{
|
|
"epoch": 2.9952,
|
|
"grad_norm": 0.2782015930914342,
|
|
"learning_rate": 1.6527036446661396e-05,
|
|
"loss": 0.3425,
|
|
"step": 234
|
|
},
|
|
{
|
|
"epoch": 3.008,
|
|
"grad_norm": 0.31433339064037064,
|
|
"learning_rate": 1.635088931591775e-05,
|
|
"loss": 0.312,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 3.0208,
|
|
"grad_norm": 0.31630189155401184,
|
|
"learning_rate": 1.6175034512658753e-05,
|
|
"loss": 0.2867,
|
|
"step": 236
|
|
},
|
|
{
|
|
"epoch": 3.0336,
|
|
"grad_norm": 0.7008268080615292,
|
|
"learning_rate": 1.5999486124479115e-05,
|
|
"loss": 0.2916,
|
|
"step": 237
|
|
},
|
|
{
|
|
"epoch": 3.0464,
|
|
"grad_norm": 0.46688092430277195,
|
|
"learning_rate": 1.5824258214426833e-05,
|
|
"loss": 0.2841,
|
|
"step": 238
|
|
},
|
|
{
|
|
"epoch": 3.0592,
|
|
"grad_norm": 0.4372201691892296,
|
|
"learning_rate": 1.5649364819876655e-05,
|
|
"loss": 0.2853,
|
|
"step": 239
|
|
},
|
|
{
|
|
"epoch": 3.072,
|
|
"grad_norm": 0.28891039619475767,
|
|
"learning_rate": 1.547481995140556e-05,
|
|
"loss": 0.2738,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 3.0848,
|
|
"grad_norm": 0.41172928619364885,
|
|
"learning_rate": 1.5300637591670357e-05,
|
|
"loss": 0.2873,
|
|
"step": 241
|
|
},
|
|
{
|
|
"epoch": 3.0976,
|
|
"grad_norm": 0.33342033959666423,
|
|
"learning_rate": 1.5126831694287564e-05,
|
|
"loss": 0.2693,
|
|
"step": 242
|
|
},
|
|
{
|
|
"epoch": 3.1104,
|
|
"grad_norm": 0.3409937922538487,
|
|
"learning_rate": 1.4953416182715566e-05,
|
|
"loss": 0.2759,
|
|
"step": 243
|
|
},
|
|
{
|
|
"epoch": 3.1232,
|
|
"grad_norm": 0.320100558341601,
|
|
"learning_rate": 1.478040494913926e-05,
|
|
"loss": 0.2836,
|
|
"step": 244
|
|
},
|
|
{
|
|
"epoch": 3.136,
|
|
"grad_norm": 0.29809795708940967,
|
|
"learning_rate": 1.460781185335713e-05,
|
|
"loss": 0.2819,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 3.1488,
|
|
"grad_norm": 0.29250467761203774,
|
|
"learning_rate": 1.443565072167095e-05,
|
|
"loss": 0.2752,
|
|
"step": 246
|
|
},
|
|
{
|
|
"epoch": 3.1616,
|
|
"grad_norm": 0.28423813424974437,
|
|
"learning_rate": 1.4263935345778202e-05,
|
|
"loss": 0.2796,
|
|
"step": 247
|
|
},
|
|
{
|
|
"epoch": 3.1744,
|
|
"grad_norm": 0.29305956201911165,
|
|
"learning_rate": 1.409267948166718e-05,
|
|
"loss": 0.2891,
|
|
"step": 248
|
|
},
|
|
{
|
|
"epoch": 3.1872,
|
|
"grad_norm": 0.29150469848488814,
|
|
"learning_rate": 1.3921896848515064e-05,
|
|
"loss": 0.2909,
|
|
"step": 249
|
|
},
|
|
{
|
|
"epoch": 3.2,
|
|
"grad_norm": 0.27886889202149356,
|
|
"learning_rate": 1.3751601127588849e-05,
|
|
"loss": 0.2779,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 3.2128,
|
|
"grad_norm": 0.2755947418181061,
|
|
"learning_rate": 1.3581805961149371e-05,
|
|
"loss": 0.2724,
|
|
"step": 251
|
|
},
|
|
{
|
|
"epoch": 3.2256,
|
|
"grad_norm": 0.2673816477671389,
|
|
"learning_rate": 1.341252495135841e-05,
|
|
"loss": 0.2861,
|
|
"step": 252
|
|
},
|
|
{
|
|
"epoch": 3.2384,
|
|
"grad_norm": 0.2635382764399931,
|
|
"learning_rate": 1.324377165918906e-05,
|
|
"loss": 0.2749,
|
|
"step": 253
|
|
},
|
|
{
|
|
"epoch": 3.2512,
|
|
"grad_norm": 0.2566074019105406,
|
|
"learning_rate": 1.3075559603339354e-05,
|
|
"loss": 0.2798,
|
|
"step": 254
|
|
},
|
|
{
|
|
"epoch": 3.2640000000000002,
|
|
"grad_norm": 0.26266710866870446,
|
|
"learning_rate": 1.2907902259149287e-05,
|
|
"loss": 0.2761,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 3.2768,
|
|
"grad_norm": 0.2805453653484367,
|
|
"learning_rate": 1.274081305752135e-05,
|
|
"loss": 0.2813,
|
|
"step": 256
|
|
},
|
|
{
|
|
"epoch": 3.2896,
|
|
"grad_norm": 0.2582997259789758,
|
|
"learning_rate": 1.2574305383844528e-05,
|
|
"loss": 0.2861,
|
|
"step": 257
|
|
},
|
|
{
|
|
"epoch": 3.3024,
|
|
"grad_norm": 0.2711321120184753,
|
|
"learning_rate": 1.2408392576922075e-05,
|
|
"loss": 0.2769,
|
|
"step": 258
|
|
},
|
|
{
|
|
"epoch": 3.3152,
|
|
"grad_norm": 0.26481288833671507,
|
|
"learning_rate": 1.2243087927902905e-05,
|
|
"loss": 0.2824,
|
|
"step": 259
|
|
},
|
|
{
|
|
"epoch": 3.328,
|
|
"grad_norm": 0.28066151058647587,
|
|
"learning_rate": 1.2078404679216864e-05,
|
|
"loss": 0.2847,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 3.3407999999999998,
|
|
"grad_norm": 0.27305714376519247,
|
|
"learning_rate": 1.1914356023513904e-05,
|
|
"loss": 0.2777,
|
|
"step": 261
|
|
},
|
|
{
|
|
"epoch": 3.3536,
|
|
"grad_norm": 0.2559955512720122,
|
|
"learning_rate": 1.1750955102607193e-05,
|
|
"loss": 0.2719,
|
|
"step": 262
|
|
},
|
|
{
|
|
"epoch": 3.3664,
|
|
"grad_norm": 0.24283413587059144,
|
|
"learning_rate": 1.1588215006420374e-05,
|
|
"loss": 0.2788,
|
|
"step": 263
|
|
},
|
|
{
|
|
"epoch": 3.3792,
|
|
"grad_norm": 0.2502212680798441,
|
|
"learning_rate": 1.1426148771938915e-05,
|
|
"loss": 0.2746,
|
|
"step": 264
|
|
},
|
|
{
|
|
"epoch": 3.392,
|
|
"grad_norm": 0.26092754422583947,
|
|
"learning_rate": 1.1264769382165748e-05,
|
|
"loss": 0.2816,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 3.4048,
|
|
"grad_norm": 0.2282597423607946,
|
|
"learning_rate": 1.110408976508118e-05,
|
|
"loss": 0.2819,
|
|
"step": 266
|
|
},
|
|
{
|
|
"epoch": 3.4176,
|
|
"grad_norm": 0.2711529173262311,
|
|
"learning_rate": 1.094412279260726e-05,
|
|
"loss": 0.2756,
|
|
"step": 267
|
|
},
|
|
{
|
|
"epoch": 3.4304,
|
|
"grad_norm": 0.23736304306624106,
|
|
"learning_rate": 1.0784881279576635e-05,
|
|
"loss": 0.2689,
|
|
"step": 268
|
|
},
|
|
{
|
|
"epoch": 3.4432,
|
|
"grad_norm": 0.26229924143448996,
|
|
"learning_rate": 1.0626377982705929e-05,
|
|
"loss": 0.2807,
|
|
"step": 269
|
|
},
|
|
{
|
|
"epoch": 3.456,
|
|
"grad_norm": 0.24160423907741094,
|
|
"learning_rate": 1.0468625599573842e-05,
|
|
"loss": 0.2849,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 3.4688,
|
|
"grad_norm": 0.24781523917685258,
|
|
"learning_rate": 1.0311636767603952e-05,
|
|
"loss": 0.2825,
|
|
"step": 271
|
|
},
|
|
{
|
|
"epoch": 3.4816,
|
|
"grad_norm": 0.23837646582381714,
|
|
"learning_rate": 1.0155424063052306e-05,
|
|
"loss": 0.2911,
|
|
"step": 272
|
|
},
|
|
{
|
|
"epoch": 3.4944,
|
|
"grad_norm": 0.22393081371619156,
|
|
"learning_rate": 1.0000000000000006e-05,
|
|
"loss": 0.2897,
|
|
"step": 273
|
|
},
|
|
{
|
|
"epoch": 3.5072,
|
|
"grad_norm": 0.2429111872222838,
|
|
"learning_rate": 9.84537702935065e-06,
|
|
"loss": 0.281,
|
|
"step": 274
|
|
},
|
|
{
|
|
"epoch": 3.52,
|
|
"grad_norm": 0.2254481115341149,
|
|
"learning_rate": 9.691567537832964e-06,
|
|
"loss": 0.2905,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 3.5328,
|
|
"grad_norm": 0.2349724201722084,
|
|
"learning_rate": 9.538583847008452e-06,
|
|
"loss": 0.2971,
|
|
"step": 276
|
|
},
|
|
{
|
|
"epoch": 3.5456,
|
|
"grad_norm": 0.23371285739416045,
|
|
"learning_rate": 9.386438212284372e-06,
|
|
"loss": 0.2777,
|
|
"step": 277
|
|
},
|
|
{
|
|
"epoch": 3.5584,
|
|
"grad_norm": 0.23792960032240412,
|
|
"learning_rate": 9.235142821931928e-06,
|
|
"loss": 0.2789,
|
|
"step": 278
|
|
},
|
|
{
|
|
"epoch": 3.5712,
|
|
"grad_norm": 0.22122647180575977,
|
|
"learning_rate": 9.084709796109907e-06,
|
|
"loss": 0.2843,
|
|
"step": 279
|
|
},
|
|
{
|
|
"epoch": 3.584,
|
|
"grad_norm": 0.2438872273005057,
|
|
"learning_rate": 8.93515118589373e-06,
|
|
"loss": 0.2822,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 3.5968,
|
|
"grad_norm": 0.23342124760266283,
|
|
"learning_rate": 8.786478972310023e-06,
|
|
"loss": 0.288,
|
|
"step": 281
|
|
},
|
|
{
|
|
"epoch": 3.6096,
|
|
"grad_norm": 0.23582335353818698,
|
|
"learning_rate": 8.638705065376887e-06,
|
|
"loss": 0.2787,
|
|
"step": 282
|
|
},
|
|
{
|
|
"epoch": 3.6224,
|
|
"grad_norm": 0.24209907403389272,
|
|
"learning_rate": 8.491841303149728e-06,
|
|
"loss": 0.2878,
|
|
"step": 283
|
|
},
|
|
{
|
|
"epoch": 3.6352,
|
|
"grad_norm": 0.21384750896331084,
|
|
"learning_rate": 8.345899450772975e-06,
|
|
"loss": 0.2897,
|
|
"step": 284
|
|
},
|
|
{
|
|
"epoch": 3.648,
|
|
"grad_norm": 0.2195417663551771,
|
|
"learning_rate": 8.200891199537549e-06,
|
|
"loss": 0.2726,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 3.6608,
|
|
"grad_norm": 0.2231301243500821,
|
|
"learning_rate": 8.056828165944282e-06,
|
|
"loss": 0.2833,
|
|
"step": 286
|
|
},
|
|
{
|
|
"epoch": 3.6736,
|
|
"grad_norm": 0.2149364649666386,
|
|
"learning_rate": 7.913721890773354e-06,
|
|
"loss": 0.2811,
|
|
"step": 287
|
|
},
|
|
{
|
|
"epoch": 3.6864,
|
|
"grad_norm": 0.21357423713236115,
|
|
"learning_rate": 7.771583838159756e-06,
|
|
"loss": 0.2826,
|
|
"step": 288
|
|
},
|
|
{
|
|
"epoch": 3.6992000000000003,
|
|
"grad_norm": 0.24825207098701313,
|
|
"learning_rate": 7.630425394674903e-06,
|
|
"loss": 0.2826,
|
|
"step": 289
|
|
},
|
|
{
|
|
"epoch": 3.7119999999999997,
|
|
"grad_norm": 0.21415212818461488,
|
|
"learning_rate": 7.49025786841445e-06,
|
|
"loss": 0.2853,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 3.7248,
|
|
"grad_norm": 0.22015828134494975,
|
|
"learning_rate": 7.3510924880924575e-06,
|
|
"loss": 0.2841,
|
|
"step": 291
|
|
},
|
|
{
|
|
"epoch": 3.7376,
|
|
"grad_norm": 0.22224706450011542,
|
|
"learning_rate": 7.212940402141808e-06,
|
|
"loss": 0.2835,
|
|
"step": 292
|
|
},
|
|
{
|
|
"epoch": 3.7504,
|
|
"grad_norm": 0.21581961016679654,
|
|
"learning_rate": 7.075812677821145e-06,
|
|
"loss": 0.2843,
|
|
"step": 293
|
|
},
|
|
{
|
|
"epoch": 3.7632,
|
|
"grad_norm": 0.21835107436796022,
|
|
"learning_rate": 6.939720300328303e-06,
|
|
"loss": 0.2688,
|
|
"step": 294
|
|
},
|
|
{
|
|
"epoch": 3.776,
|
|
"grad_norm": 0.212157439367985,
|
|
"learning_rate": 6.8046741719202385e-06,
|
|
"loss": 0.2754,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 3.7888,
|
|
"grad_norm": 0.21531406072877302,
|
|
"learning_rate": 6.67068511103971e-06,
|
|
"loss": 0.28,
|
|
"step": 296
|
|
},
|
|
{
|
|
"epoch": 3.8016,
|
|
"grad_norm": 0.2334993523584025,
|
|
"learning_rate": 6.537763851448593e-06,
|
|
"loss": 0.2797,
|
|
"step": 297
|
|
},
|
|
{
|
|
"epoch": 3.8144,
|
|
"grad_norm": 0.21250002898841552,
|
|
"learning_rate": 6.4059210413680175e-06,
|
|
"loss": 0.2783,
|
|
"step": 298
|
|
},
|
|
{
|
|
"epoch": 3.8272,
|
|
"grad_norm": 0.2050927549154722,
|
|
"learning_rate": 6.275167242625331e-06,
|
|
"loss": 0.2806,
|
|
"step": 299
|
|
},
|
|
{
|
|
"epoch": 3.84,
|
|
"grad_norm": 0.23528957358515223,
|
|
"learning_rate": 6.145512929808013e-06,
|
|
"loss": 0.2823,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 3.8528000000000002,
|
|
"grad_norm": 0.21866030120409596,
|
|
"learning_rate": 6.016968489424572e-06,
|
|
"loss": 0.2694,
|
|
"step": 301
|
|
},
|
|
{
|
|
"epoch": 3.8656,
|
|
"grad_norm": 0.1986886641313073,
|
|
"learning_rate": 5.889544219072465e-06,
|
|
"loss": 0.2815,
|
|
"step": 302
|
|
},
|
|
{
|
|
"epoch": 3.8784,
|
|
"grad_norm": 0.20407358585815805,
|
|
"learning_rate": 5.7632503266131925e-06,
|
|
"loss": 0.281,
|
|
"step": 303
|
|
},
|
|
{
|
|
"epoch": 3.8912,
|
|
"grad_norm": 0.21795566932605281,
|
|
"learning_rate": 5.638096929354522e-06,
|
|
"loss": 0.2698,
|
|
"step": 304
|
|
},
|
|
{
|
|
"epoch": 3.904,
|
|
"grad_norm": 0.20828825849742277,
|
|
"learning_rate": 5.514094053240035e-06,
|
|
"loss": 0.2814,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 3.9168,
|
|
"grad_norm": 0.20552858430321327,
|
|
"learning_rate": 5.39125163204594e-06,
|
|
"loss": 0.281,
|
|
"step": 306
|
|
},
|
|
{
|
|
"epoch": 3.9295999999999998,
|
|
"grad_norm": 0.18563231992381732,
|
|
"learning_rate": 5.269579506585259e-06,
|
|
"loss": 0.2784,
|
|
"step": 307
|
|
},
|
|
{
|
|
"epoch": 3.9424,
|
|
"grad_norm": 0.2181494958934182,
|
|
"learning_rate": 5.149087423919541e-06,
|
|
"loss": 0.2684,
|
|
"step": 308
|
|
},
|
|
{
|
|
"epoch": 3.9552,
|
|
"grad_norm": 0.20575475264490437,
|
|
"learning_rate": 5.029785036577976e-06,
|
|
"loss": 0.2812,
|
|
"step": 309
|
|
},
|
|
{
|
|
"epoch": 3.968,
|
|
"grad_norm": 0.20762835949111869,
|
|
"learning_rate": 4.911681901784198e-06,
|
|
"loss": 0.2717,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 3.9808,
|
|
"grad_norm": 0.19160088442220838,
|
|
"learning_rate": 4.794787480690597e-06,
|
|
"loss": 0.2738,
|
|
"step": 311
|
|
},
|
|
{
|
|
"epoch": 3.9936,
|
|
"grad_norm": 0.2024203895067496,
|
|
"learning_rate": 4.679111137620442e-06,
|
|
"loss": 0.2782,
|
|
"step": 312
|
|
},
|
|
{
|
|
"epoch": 4.0064,
|
|
"grad_norm": 0.2483219477733269,
|
|
"learning_rate": 4.5646621393177e-06,
|
|
"loss": 0.2671,
|
|
"step": 313
|
|
},
|
|
{
|
|
"epoch": 4.0192,
|
|
"grad_norm": 0.31427588349183117,
|
|
"learning_rate": 4.451449654204685e-06,
|
|
"loss": 0.2487,
|
|
"step": 314
|
|
},
|
|
{
|
|
"epoch": 4.032,
|
|
"grad_norm": 0.2210039512386439,
|
|
"learning_rate": 4.339482751647557e-06,
|
|
"loss": 0.2463,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 4.0448,
|
|
"grad_norm": 0.26443806180928325,
|
|
"learning_rate": 4.228770401229824e-06,
|
|
"loss": 0.2385,
|
|
"step": 316
|
|
},
|
|
{
|
|
"epoch": 4.0576,
|
|
"grad_norm": 0.3912919431430795,
|
|
"learning_rate": 4.119321472033779e-06,
|
|
"loss": 0.2463,
|
|
"step": 317
|
|
},
|
|
{
|
|
"epoch": 4.0704,
|
|
"grad_norm": 0.25929046643811793,
|
|
"learning_rate": 4.011144731929981e-06,
|
|
"loss": 0.2441,
|
|
"step": 318
|
|
},
|
|
{
|
|
"epoch": 4.0832,
|
|
"grad_norm": 0.24349599096142088,
|
|
"learning_rate": 3.904248846874894e-06,
|
|
"loss": 0.2356,
|
|
"step": 319
|
|
},
|
|
{
|
|
"epoch": 4.096,
|
|
"grad_norm": 0.2900612365803001,
|
|
"learning_rate": 3.7986423802166705e-06,
|
|
"loss": 0.2515,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 4.1088,
|
|
"grad_norm": 0.2873414105221977,
|
|
"learning_rate": 3.694333792009115e-06,
|
|
"loss": 0.2536,
|
|
"step": 321
|
|
},
|
|
{
|
|
"epoch": 4.1216,
|
|
"grad_norm": 0.23751942576573712,
|
|
"learning_rate": 3.5913314383339937e-06,
|
|
"loss": 0.2503,
|
|
"step": 322
|
|
},
|
|
{
|
|
"epoch": 4.1344,
|
|
"grad_norm": 0.21072311310845263,
|
|
"learning_rate": 3.4896435706316e-06,
|
|
"loss": 0.2481,
|
|
"step": 323
|
|
},
|
|
{
|
|
"epoch": 4.1472,
|
|
"grad_norm": 0.2434813470084433,
|
|
"learning_rate": 3.3892783350397675e-06,
|
|
"loss": 0.234,
|
|
"step": 324
|
|
},
|
|
{
|
|
"epoch": 4.16,
|
|
"grad_norm": 0.268913173192014,
|
|
"learning_rate": 3.290243771741275e-06,
|
|
"loss": 0.2427,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 4.1728,
|
|
"grad_norm": 0.2518464525686985,
|
|
"learning_rate": 3.1925478143197418e-06,
|
|
"loss": 0.2453,
|
|
"step": 326
|
|
},
|
|
{
|
|
"epoch": 4.1856,
|
|
"grad_norm": 0.2202620097200464,
|
|
"learning_rate": 3.0961982891241083e-06,
|
|
"loss": 0.2476,
|
|
"step": 327
|
|
},
|
|
{
|
|
"epoch": 4.1984,
|
|
"grad_norm": 0.20611433865734524,
|
|
"learning_rate": 3.001202914641628e-06,
|
|
"loss": 0.2529,
|
|
"step": 328
|
|
},
|
|
{
|
|
"epoch": 4.2112,
|
|
"grad_norm": 0.23978958306869538,
|
|
"learning_rate": 2.907569300879596e-06,
|
|
"loss": 0.2365,
|
|
"step": 329
|
|
},
|
|
{
|
|
"epoch": 4.224,
|
|
"grad_norm": 0.24793303268384181,
|
|
"learning_rate": 2.815304948755664e-06,
|
|
"loss": 0.245,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 4.2368,
|
|
"grad_norm": 0.2211020129342562,
|
|
"learning_rate": 2.7244172494969978e-06,
|
|
"loss": 0.2228,
|
|
"step": 331
|
|
},
|
|
{
|
|
"epoch": 4.2496,
|
|
"grad_norm": 0.21422382276077329,
|
|
"learning_rate": 2.6349134840481294e-06,
|
|
"loss": 0.2534,
|
|
"step": 332
|
|
},
|
|
{
|
|
"epoch": 4.2624,
|
|
"grad_norm": 0.23353733968476206,
|
|
"learning_rate": 2.546800822487714e-06,
|
|
"loss": 0.2333,
|
|
"step": 333
|
|
},
|
|
{
|
|
"epoch": 4.2752,
|
|
"grad_norm": 0.22380233207256295,
|
|
"learning_rate": 2.4600863234541338e-06,
|
|
"loss": 0.2506,
|
|
"step": 334
|
|
},
|
|
{
|
|
"epoch": 4.288,
|
|
"grad_norm": 0.22667996547433356,
|
|
"learning_rate": 2.374776933580025e-06,
|
|
"loss": 0.2418,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 4.3008,
|
|
"grad_norm": 0.2083067132416159,
|
|
"learning_rate": 2.2908794869358044e-06,
|
|
"loss": 0.2348,
|
|
"step": 336
|
|
},
|
|
{
|
|
"epoch": 4.3136,
|
|
"grad_norm": 0.1899808874315056,
|
|
"learning_rate": 2.2084007044821764e-06,
|
|
"loss": 0.2472,
|
|
"step": 337
|
|
},
|
|
{
|
|
"epoch": 4.3264,
|
|
"grad_norm": 0.19416358137566275,
|
|
"learning_rate": 2.127347193531757e-06,
|
|
"loss": 0.2399,
|
|
"step": 338
|
|
},
|
|
{
|
|
"epoch": 4.3392,
|
|
"grad_norm": 0.19459645408383455,
|
|
"learning_rate": 2.0477254472197237e-06,
|
|
"loss": 0.2273,
|
|
"step": 339
|
|
},
|
|
{
|
|
"epoch": 4.352,
|
|
"grad_norm": 0.18919120751562665,
|
|
"learning_rate": 1.96954184398368e-06,
|
|
"loss": 0.2388,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 4.3648,
|
|
"grad_norm": 0.19581399230843038,
|
|
"learning_rate": 1.8928026470526917e-06,
|
|
"loss": 0.2371,
|
|
"step": 341
|
|
},
|
|
{
|
|
"epoch": 4.3776,
|
|
"grad_norm": 0.19539124986378142,
|
|
"learning_rate": 1.817514003945524e-06,
|
|
"loss": 0.24,
|
|
"step": 342
|
|
},
|
|
{
|
|
"epoch": 4.3904,
|
|
"grad_norm": 0.18798319840625727,
|
|
"learning_rate": 1.743681945978184e-06,
|
|
"loss": 0.2425,
|
|
"step": 343
|
|
},
|
|
{
|
|
"epoch": 4.4032,
|
|
"grad_norm": 0.19148378298433658,
|
|
"learning_rate": 1.6713123877807413e-06,
|
|
"loss": 0.2355,
|
|
"step": 344
|
|
},
|
|
{
|
|
"epoch": 4.416,
|
|
"grad_norm": 0.19209574289693548,
|
|
"learning_rate": 1.6004111268235156e-06,
|
|
"loss": 0.2399,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 4.4288,
|
|
"grad_norm": 0.19840558163677857,
|
|
"learning_rate": 1.5309838429526714e-06,
|
|
"loss": 0.2421,
|
|
"step": 346
|
|
},
|
|
{
|
|
"epoch": 4.4416,
|
|
"grad_norm": 0.1806706307524296,
|
|
"learning_rate": 1.4630360979351644e-06,
|
|
"loss": 0.2324,
|
|
"step": 347
|
|
},
|
|
{
|
|
"epoch": 4.4544,
|
|
"grad_norm": 0.17851900712531546,
|
|
"learning_rate": 1.396573335013236e-06,
|
|
"loss": 0.2296,
|
|
"step": 348
|
|
},
|
|
{
|
|
"epoch": 4.4672,
|
|
"grad_norm": 0.1851790531424452,
|
|
"learning_rate": 1.3316008784683265e-06,
|
|
"loss": 0.2455,
|
|
"step": 349
|
|
},
|
|
{
|
|
"epoch": 4.48,
|
|
"grad_norm": 0.18420836933280219,
|
|
"learning_rate": 1.2681239331945695e-06,
|
|
"loss": 0.2392,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 4.4928,
|
|
"grad_norm": 0.182888054979642,
|
|
"learning_rate": 1.2061475842818337e-06,
|
|
"loss": 0.2372,
|
|
"step": 351
|
|
},
|
|
{
|
|
"epoch": 4.5056,
|
|
"grad_norm": 0.18199196177578558,
|
|
"learning_rate": 1.1456767966083393e-06,
|
|
"loss": 0.2429,
|
|
"step": 352
|
|
},
|
|
{
|
|
"epoch": 4.5184,
|
|
"grad_norm": 0.18285968384881784,
|
|
"learning_rate": 1.086716414442952e-06,
|
|
"loss": 0.2424,
|
|
"step": 353
|
|
},
|
|
{
|
|
"epoch": 4.5312,
|
|
"grad_norm": 0.18120132437676076,
|
|
"learning_rate": 1.0292711610570904e-06,
|
|
"loss": 0.2405,
|
|
"step": 354
|
|
},
|
|
{
|
|
"epoch": 4.5440000000000005,
|
|
"grad_norm": 0.18256647144707736,
|
|
"learning_rate": 9.733456383463658e-07,
|
|
"loss": 0.2403,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 4.5568,
|
|
"grad_norm": 0.1806229959243068,
|
|
"learning_rate": 9.189443264619102e-07,
|
|
"loss": 0.2358,
|
|
"step": 356
|
|
},
|
|
{
|
|
"epoch": 4.5696,
|
|
"grad_norm": 0.17903500347908813,
|
|
"learning_rate": 8.660715834514977e-07,
|
|
"loss": 0.242,
|
|
"step": 357
|
|
},
|
|
{
|
|
"epoch": 4.5824,
|
|
"grad_norm": 0.18322902020731868,
|
|
"learning_rate": 8.147316449103959e-07,
|
|
"loss": 0.2353,
|
|
"step": 358
|
|
},
|
|
{
|
|
"epoch": 4.5952,
|
|
"grad_norm": 0.18920089983891739,
|
|
"learning_rate": 7.649286236420806e-07,
|
|
"loss": 0.2411,
|
|
"step": 359
|
|
},
|
|
{
|
|
"epoch": 4.608,
|
|
"grad_norm": 0.1715008809199515,
|
|
"learning_rate": 7.166665093287539e-07,
|
|
"loss": 0.2324,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 4.6208,
|
|
"grad_norm": 0.18710310075634906,
|
|
"learning_rate": 6.69949168211721e-07,
|
|
"loss": 0.2437,
|
|
"step": 361
|
|
},
|
|
{
|
|
"epoch": 4.6336,
|
|
"grad_norm": 0.18506851044805525,
|
|
"learning_rate": 6.247803427816945e-07,
|
|
"loss": 0.2455,
|
|
"step": 362
|
|
},
|
|
{
|
|
"epoch": 4.6464,
|
|
"grad_norm": 0.18094845128973183,
|
|
"learning_rate": 5.811636514789598e-07,
|
|
"loss": 0.2258,
|
|
"step": 363
|
|
},
|
|
{
|
|
"epoch": 4.6592,
|
|
"grad_norm": 0.17435332481727803,
|
|
"learning_rate": 5.391025884035239e-07,
|
|
"loss": 0.239,
|
|
"step": 364
|
|
},
|
|
{
|
|
"epoch": 4.672,
|
|
"grad_norm": 0.17695486262206614,
|
|
"learning_rate": 4.986005230351954e-07,
|
|
"loss": 0.2269,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 4.6848,
|
|
"grad_norm": 0.18275683820670355,
|
|
"learning_rate": 4.5966069996365993e-07,
|
|
"loss": 0.2508,
|
|
"step": 366
|
|
},
|
|
{
|
|
"epoch": 4.6975999999999996,
|
|
"grad_norm": 0.1837979328938096,
|
|
"learning_rate": 4.22286238628562e-07,
|
|
"loss": 0.2368,
|
|
"step": 367
|
|
},
|
|
{
|
|
"epoch": 4.7104,
|
|
"grad_norm": 0.18220151570131113,
|
|
"learning_rate": 3.8648013306960664e-07,
|
|
"loss": 0.2477,
|
|
"step": 368
|
|
},
|
|
{
|
|
"epoch": 4.7232,
|
|
"grad_norm": 0.1730991293089666,
|
|
"learning_rate": 3.522452516867048e-07,
|
|
"loss": 0.2469,
|
|
"step": 369
|
|
},
|
|
{
|
|
"epoch": 4.736,
|
|
"grad_norm": 0.19236348216935706,
|
|
"learning_rate": 3.1958433701019697e-07,
|
|
"loss": 0.2421,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 4.7488,
|
|
"grad_norm": 0.1706653957534717,
|
|
"learning_rate": 2.8850000548115155e-07,
|
|
"loss": 0.2441,
|
|
"step": 371
|
|
},
|
|
{
|
|
"epoch": 4.7616,
|
|
"grad_norm": 0.18911082173661192,
|
|
"learning_rate": 2.5899474724174313e-07,
|
|
"loss": 0.2423,
|
|
"step": 372
|
|
},
|
|
{
|
|
"epoch": 4.7744,
|
|
"grad_norm": 0.17618675679611007,
|
|
"learning_rate": 2.3107092593579905e-07,
|
|
"loss": 0.259,
|
|
"step": 373
|
|
},
|
|
{
|
|
"epoch": 4.7872,
|
|
"grad_norm": 0.17961036651229495,
|
|
"learning_rate": 2.0473077851942858e-07,
|
|
"loss": 0.2478,
|
|
"step": 374
|
|
},
|
|
{
|
|
"epoch": 4.8,
|
|
"grad_norm": 0.17730836222535062,
|
|
"learning_rate": 1.799764150818306e-07,
|
|
"loss": 0.2484,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 4.8128,
|
|
"grad_norm": 0.17676500581389382,
|
|
"learning_rate": 1.5680981867625566e-07,
|
|
"loss": 0.2365,
|
|
"step": 376
|
|
},
|
|
{
|
|
"epoch": 4.8256,
|
|
"grad_norm": 0.17587484900010034,
|
|
"learning_rate": 1.3523284516113955e-07,
|
|
"loss": 0.2543,
|
|
"step": 377
|
|
},
|
|
{
|
|
"epoch": 4.8384,
|
|
"grad_norm": 0.17111307611066914,
|
|
"learning_rate": 1.1524722305144231e-07,
|
|
"loss": 0.2449,
|
|
"step": 378
|
|
},
|
|
{
|
|
"epoch": 4.8512,
|
|
"grad_norm": 0.16914105389943437,
|
|
"learning_rate": 9.685455338016347e-08,
|
|
"loss": 0.2415,
|
|
"step": 379
|
|
},
|
|
{
|
|
"epoch": 4.864,
|
|
"grad_norm": 0.17406515355425717,
|
|
"learning_rate": 8.005630957010014e-08,
|
|
"loss": 0.2429,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 4.8768,
|
|
"grad_norm": 0.17327135001293076,
|
|
"learning_rate": 6.485383731580142e-08,
|
|
"loss": 0.2505,
|
|
"step": 381
|
|
},
|
|
{
|
|
"epoch": 4.8896,
|
|
"grad_norm": 0.1762049065101459,
|
|
"learning_rate": 5.1248354475768034e-08,
|
|
"loss": 0.234,
|
|
"step": 382
|
|
},
|
|
{
|
|
"epoch": 4.9024,
|
|
"grad_norm": 0.17124756759793125,
|
|
"learning_rate": 3.924095097489922e-08,
|
|
"loss": 0.2415,
|
|
"step": 383
|
|
},
|
|
{
|
|
"epoch": 4.9152000000000005,
|
|
"grad_norm": 0.18444523509295194,
|
|
"learning_rate": 2.8832588717164766e-08,
|
|
"loss": 0.2261,
|
|
"step": 384
|
|
},
|
|
{
|
|
"epoch": 4.928,
|
|
"grad_norm": 0.18310827178900155,
|
|
"learning_rate": 2.0024101508555604e-08,
|
|
"loss": 0.2434,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 4.9408,
|
|
"grad_norm": 0.17186707504148238,
|
|
"learning_rate": 1.281619499029274e-08,
|
|
"loss": 0.2454,
|
|
"step": 386
|
|
},
|
|
{
|
|
"epoch": 4.9536,
|
|
"grad_norm": 0.17818375445377446,
|
|
"learning_rate": 7.209446582292501e-09,
|
|
"loss": 0.2519,
|
|
"step": 387
|
|
},
|
|
{
|
|
"epoch": 4.9664,
|
|
"grad_norm": 0.16976076854341357,
|
|
"learning_rate": 3.2043054369057523e-09,
|
|
"loss": 0.2388,
|
|
"step": 388
|
|
},
|
|
{
|
|
"epoch": 4.9792,
|
|
"grad_norm": 0.17562050527830075,
|
|
"learning_rate": 8.010924029533406e-10,
|
|
"loss": 0.2355,
|
|
"step": 389
|
|
},
|
|
{
|
|
"epoch": 4.992,
|
|
"grad_norm": 0.24606256837815751,
|
|
"learning_rate": 0.0,
|
|
"loss": 0.2438,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 4.992,
|
|
"step": 390,
|
|
"total_flos": 1.9952099917069025e+18,
|
|
"train_loss": 0.36405690262715024,
|
|
"train_runtime": 60425.7754,
|
|
"train_samples_per_second": 0.827,
|
|
"train_steps_per_second": 0.006
|
|
}
|
|
],
|
|
"logging_steps": 1,
|
|
"max_steps": 390,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 5,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1.9952099917069025e+18,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|