BENT-PubMedBERT-NER-Chemical / trainer_state.json
pruas's picture
Upload 11 files
11aacd0
{
"best_metric": 0.04133550822734833,
"best_model_checkpoint": "data/models/chemical/final/checkpoint-20571",
"epoch": 1.0,
"global_step": 20571,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 9.922807350153128e-05,
"loss": 0.094,
"step": 500
},
{
"epoch": 0.05,
"learning_rate": 9.675614700306256e-05,
"loss": 0.0704,
"step": 1000
},
{
"epoch": 0.07,
"learning_rate": 9.428422050459385e-05,
"loss": 0.0626,
"step": 1500
},
{
"epoch": 0.1,
"learning_rate": 9.181229400612513e-05,
"loss": 0.0621,
"step": 2000
},
{
"epoch": 0.12,
"learning_rate": 8.934036750765641e-05,
"loss": 0.0586,
"step": 2500
},
{
"epoch": 0.15,
"learning_rate": 8.68684410091877e-05,
"loss": 0.0535,
"step": 3000
},
{
"epoch": 0.17,
"learning_rate": 8.439651451071898e-05,
"loss": 0.0553,
"step": 3500
},
{
"epoch": 0.19,
"learning_rate": 8.192458801225026e-05,
"loss": 0.0514,
"step": 4000
},
{
"epoch": 0.22,
"learning_rate": 7.945266151378155e-05,
"loss": 0.0513,
"step": 4500
},
{
"epoch": 0.24,
"learning_rate": 7.698073501531283e-05,
"loss": 0.0497,
"step": 5000
},
{
"epoch": 0.27,
"learning_rate": 7.45088085168441e-05,
"loss": 0.0466,
"step": 5500
},
{
"epoch": 0.29,
"learning_rate": 7.203688201837538e-05,
"loss": 0.0481,
"step": 6000
},
{
"epoch": 0.32,
"learning_rate": 6.956495551990666e-05,
"loss": 0.0433,
"step": 6500
},
{
"epoch": 0.34,
"learning_rate": 6.709302902143795e-05,
"loss": 0.0442,
"step": 7000
},
{
"epoch": 0.36,
"learning_rate": 6.462110252296923e-05,
"loss": 0.0471,
"step": 7500
},
{
"epoch": 0.39,
"learning_rate": 6.214917602450051e-05,
"loss": 0.0451,
"step": 8000
},
{
"epoch": 0.41,
"learning_rate": 5.96772495260318e-05,
"loss": 0.0428,
"step": 8500
},
{
"epoch": 0.44,
"learning_rate": 5.720532302756308e-05,
"loss": 0.0406,
"step": 9000
},
{
"epoch": 0.46,
"learning_rate": 5.473339652909436e-05,
"loss": 0.0448,
"step": 9500
},
{
"epoch": 0.49,
"learning_rate": 5.226147003062564e-05,
"loss": 0.0413,
"step": 10000
},
{
"epoch": 0.51,
"learning_rate": 4.9789543532156925e-05,
"loss": 0.0408,
"step": 10500
},
{
"epoch": 0.53,
"learning_rate": 4.7317617033688203e-05,
"loss": 0.0386,
"step": 11000
},
{
"epoch": 0.56,
"learning_rate": 4.484569053521948e-05,
"loss": 0.0402,
"step": 11500
},
{
"epoch": 0.58,
"learning_rate": 4.2373764036750766e-05,
"loss": 0.0365,
"step": 12000
},
{
"epoch": 0.61,
"learning_rate": 3.990183753828205e-05,
"loss": 0.038,
"step": 12500
},
{
"epoch": 0.63,
"learning_rate": 3.742991103981333e-05,
"loss": 0.0346,
"step": 13000
},
{
"epoch": 0.66,
"learning_rate": 3.4957984541344614e-05,
"loss": 0.0356,
"step": 13500
},
{
"epoch": 0.68,
"learning_rate": 3.248605804287589e-05,
"loss": 0.0346,
"step": 14000
},
{
"epoch": 0.7,
"learning_rate": 3.0014131544407173e-05,
"loss": 0.0304,
"step": 14500
},
{
"epoch": 0.73,
"learning_rate": 2.754220504593846e-05,
"loss": 0.0334,
"step": 15000
},
{
"epoch": 0.75,
"learning_rate": 2.5070278547469743e-05,
"loss": 0.0312,
"step": 15500
},
{
"epoch": 0.78,
"learning_rate": 2.259835204900102e-05,
"loss": 0.0333,
"step": 16000
},
{
"epoch": 0.8,
"learning_rate": 2.0126425550532302e-05,
"loss": 0.0321,
"step": 16500
},
{
"epoch": 0.83,
"learning_rate": 1.7654499052063587e-05,
"loss": 0.0338,
"step": 17000
},
{
"epoch": 0.85,
"learning_rate": 1.5182572553594867e-05,
"loss": 0.0309,
"step": 17500
},
{
"epoch": 0.88,
"learning_rate": 1.271064605512615e-05,
"loss": 0.0295,
"step": 18000
},
{
"epoch": 0.9,
"learning_rate": 1.023871955665743e-05,
"loss": 0.0281,
"step": 18500
},
{
"epoch": 0.92,
"learning_rate": 7.766793058188713e-06,
"loss": 0.0303,
"step": 19000
},
{
"epoch": 0.95,
"learning_rate": 5.294866559719994e-06,
"loss": 0.0294,
"step": 19500
},
{
"epoch": 0.97,
"learning_rate": 2.822940061251276e-06,
"loss": 0.0274,
"step": 20000
},
{
"epoch": 1.0,
"learning_rate": 3.51013562782558e-07,
"loss": 0.0283,
"step": 20500
},
{
"epoch": 1.0,
"eval_accuracy": 0.9787234042553191,
"eval_f1": 0.9333333333333333,
"eval_loss": 0.04133550822734833,
"eval_precision": 1.0,
"eval_recall": 0.875,
"eval_runtime": 5.3542,
"eval_samples_per_second": 0.747,
"eval_steps_per_second": 0.187,
"step": 20571
}
],
"max_steps": 20571,
"num_train_epochs": 1,
"total_flos": 8.6000072575104e+16,
"trial_name": null,
"trial_params": null
}