leixa's picture
Training in progress, step 100, checkpoint
80e2206 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.8691588785046729,
"eval_steps": 9,
"global_step": 100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.018691588785046728,
"eval_loss": 1.8203117847442627,
"eval_runtime": 4.1702,
"eval_samples_per_second": 21.582,
"eval_steps_per_second": 2.878,
"step": 1
},
{
"epoch": 0.056074766355140186,
"grad_norm": 0.05738469213247299,
"learning_rate": 3e-05,
"loss": 1.8433,
"step": 3
},
{
"epoch": 0.11214953271028037,
"grad_norm": 0.06689329445362091,
"learning_rate": 6e-05,
"loss": 1.8043,
"step": 6
},
{
"epoch": 0.16822429906542055,
"grad_norm": 0.06083233654499054,
"learning_rate": 9e-05,
"loss": 1.7117,
"step": 9
},
{
"epoch": 0.16822429906542055,
"eval_loss": 1.8101048469543457,
"eval_runtime": 4.2267,
"eval_samples_per_second": 21.293,
"eval_steps_per_second": 2.839,
"step": 9
},
{
"epoch": 0.22429906542056074,
"grad_norm": 0.08025555312633514,
"learning_rate": 9.987820251299122e-05,
"loss": 1.6754,
"step": 12
},
{
"epoch": 0.2803738317757009,
"grad_norm": 0.07243169844150543,
"learning_rate": 9.924038765061042e-05,
"loss": 1.7195,
"step": 15
},
{
"epoch": 0.3364485981308411,
"grad_norm": 0.07378420233726501,
"learning_rate": 9.806308479691595e-05,
"loss": 1.6835,
"step": 18
},
{
"epoch": 0.3364485981308411,
"eval_loss": 1.7150883674621582,
"eval_runtime": 4.2453,
"eval_samples_per_second": 21.2,
"eval_steps_per_second": 2.827,
"step": 18
},
{
"epoch": 0.3925233644859813,
"grad_norm": 0.09596385061740875,
"learning_rate": 9.635919272833938e-05,
"loss": 1.7114,
"step": 21
},
{
"epoch": 0.4485981308411215,
"grad_norm": 0.06433645635843277,
"learning_rate": 9.414737964294636e-05,
"loss": 1.6425,
"step": 24
},
{
"epoch": 0.5046728971962616,
"grad_norm": 0.06536176800727844,
"learning_rate": 9.145187862775209e-05,
"loss": 1.547,
"step": 27
},
{
"epoch": 0.5046728971962616,
"eval_loss": 1.6153008937835693,
"eval_runtime": 4.2498,
"eval_samples_per_second": 21.177,
"eval_steps_per_second": 2.824,
"step": 27
},
{
"epoch": 0.5607476635514018,
"grad_norm": 0.07019589841365814,
"learning_rate": 8.83022221559489e-05,
"loss": 1.6491,
"step": 30
},
{
"epoch": 0.616822429906542,
"grad_norm": 0.07840248197317123,
"learning_rate": 8.473291852294987e-05,
"loss": 1.8556,
"step": 33
},
{
"epoch": 0.6728971962616822,
"grad_norm": 0.060799770057201385,
"learning_rate": 8.07830737662829e-05,
"loss": 1.3197,
"step": 36
},
{
"epoch": 0.6728971962616822,
"eval_loss": 1.5317968130111694,
"eval_runtime": 4.249,
"eval_samples_per_second": 21.181,
"eval_steps_per_second": 2.824,
"step": 36
},
{
"epoch": 0.7289719626168224,
"grad_norm": 0.0651288628578186,
"learning_rate": 7.649596321166024e-05,
"loss": 1.6686,
"step": 39
},
{
"epoch": 0.7850467289719626,
"grad_norm": 0.07936910539865494,
"learning_rate": 7.191855733945387e-05,
"loss": 1.4417,
"step": 42
},
{
"epoch": 0.8411214953271028,
"grad_norm": 0.057609252631664276,
"learning_rate": 6.710100716628344e-05,
"loss": 1.4829,
"step": 45
},
{
"epoch": 0.8411214953271028,
"eval_loss": 1.4880197048187256,
"eval_runtime": 4.2472,
"eval_samples_per_second": 21.19,
"eval_steps_per_second": 2.825,
"step": 45
},
{
"epoch": 0.897196261682243,
"grad_norm": 0.06338804960250854,
"learning_rate": 6.209609477998338e-05,
"loss": 1.6396,
"step": 48
},
{
"epoch": 0.9532710280373832,
"grad_norm": 0.05713721737265587,
"learning_rate": 5.695865504800327e-05,
"loss": 1.5775,
"step": 51
},
{
"epoch": 1.0093457943925233,
"grad_norm": 0.11755422502756119,
"learning_rate": 5.174497483512506e-05,
"loss": 1.8238,
"step": 54
},
{
"epoch": 1.0093457943925233,
"eval_loss": 1.466562032699585,
"eval_runtime": 4.2554,
"eval_samples_per_second": 21.15,
"eval_steps_per_second": 2.82,
"step": 54
},
{
"epoch": 1.0654205607476634,
"grad_norm": 0.061317555606365204,
"learning_rate": 4.6512176312793736e-05,
"loss": 1.5379,
"step": 57
},
{
"epoch": 1.1214953271028036,
"grad_norm": 0.05398834869265556,
"learning_rate": 4.131759111665349e-05,
"loss": 1.484,
"step": 60
},
{
"epoch": 1.1775700934579438,
"grad_norm": 0.06205625832080841,
"learning_rate": 3.6218132209150045e-05,
"loss": 1.4773,
"step": 63
},
{
"epoch": 1.1775700934579438,
"eval_loss": 1.4554001092910767,
"eval_runtime": 4.2536,
"eval_samples_per_second": 21.159,
"eval_steps_per_second": 2.821,
"step": 63
},
{
"epoch": 1.233644859813084,
"grad_norm": 0.05923362821340561,
"learning_rate": 3.12696703292044e-05,
"loss": 1.3561,
"step": 66
},
{
"epoch": 1.2897196261682242,
"grad_norm": 0.05100737139582634,
"learning_rate": 2.6526421860705473e-05,
"loss": 1.4121,
"step": 69
},
{
"epoch": 1.3457943925233644,
"grad_norm": 0.05129067227244377,
"learning_rate": 2.2040354826462668e-05,
"loss": 1.4624,
"step": 72
},
{
"epoch": 1.3457943925233644,
"eval_loss": 1.449897050857544,
"eval_runtime": 4.2543,
"eval_samples_per_second": 21.155,
"eval_steps_per_second": 2.821,
"step": 72
},
{
"epoch": 1.4018691588785046,
"grad_norm": 0.06185724586248398,
"learning_rate": 1.7860619515673033e-05,
"loss": 1.5703,
"step": 75
},
{
"epoch": 1.4579439252336448,
"grad_norm": 0.05341401696205139,
"learning_rate": 1.4033009983067452e-05,
"loss": 1.5141,
"step": 78
},
{
"epoch": 1.514018691588785,
"grad_norm": 0.05680753290653229,
"learning_rate": 1.0599462319663905e-05,
"loss": 1.6556,
"step": 81
},
{
"epoch": 1.514018691588785,
"eval_loss": 1.4462116956710815,
"eval_runtime": 4.2559,
"eval_samples_per_second": 21.147,
"eval_steps_per_second": 2.82,
"step": 81
},
{
"epoch": 1.5700934579439252,
"grad_norm": 0.05871470645070076,
"learning_rate": 7.597595192178702e-06,
"loss": 1.4615,
"step": 84
},
{
"epoch": 1.6261682242990654,
"grad_norm": 0.05626693740487099,
"learning_rate": 5.060297685041659e-06,
"loss": 1.38,
"step": 87
},
{
"epoch": 1.6822429906542056,
"grad_norm": 0.05646049231290817,
"learning_rate": 3.0153689607045845e-06,
"loss": 1.4491,
"step": 90
},
{
"epoch": 1.6822429906542056,
"eval_loss": 1.4449619054794312,
"eval_runtime": 4.2532,
"eval_samples_per_second": 21.161,
"eval_steps_per_second": 2.821,
"step": 90
},
{
"epoch": 1.7383177570093458,
"grad_norm": 0.056759607046842575,
"learning_rate": 1.4852136862001764e-06,
"loss": 1.6806,
"step": 93
},
{
"epoch": 1.794392523364486,
"grad_norm": 0.06771969795227051,
"learning_rate": 4.865965629214819e-07,
"loss": 1.4989,
"step": 96
},
{
"epoch": 1.8504672897196262,
"grad_norm": 0.04481120780110359,
"learning_rate": 3.04586490452119e-08,
"loss": 1.3905,
"step": 99
},
{
"epoch": 1.8504672897196262,
"eval_loss": 1.4447249174118042,
"eval_runtime": 4.2552,
"eval_samples_per_second": 21.151,
"eval_steps_per_second": 2.82,
"step": 99
}
],
"logging_steps": 3,
"max_steps": 100,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 9,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.642145898482893e+16,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}