|
{ |
|
"best_metric": 0.16890692710876465, |
|
"best_model_checkpoint": "./indic_ur/checkpoint-6250", |
|
"epoch": 5.0, |
|
"global_step": 6250, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 1.885714285714286e-05, |
|
"loss": 0.8064, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"learning_rate": 1.7714285714285717e-05, |
|
"loss": 0.5114, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_LOC_f1": 0.8276643990929706, |
|
"eval_ORG_f1": 0.6985534297713485, |
|
"eval_PER_f1": 0.6433615819209039, |
|
"eval_loss": 0.41389304399490356, |
|
"eval_overall_accuracy": 0.8477015558698727, |
|
"eval_overall_f1": 0.7343858431644692, |
|
"eval_overall_precision": 0.7192863211554801, |
|
"eval_overall_recall": 0.7501329080276449, |
|
"eval_runtime": 7.4875, |
|
"eval_samples_per_second": 133.557, |
|
"eval_steps_per_second": 8.414, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"learning_rate": 1.6571428571428574e-05, |
|
"loss": 0.4147, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"learning_rate": 1.542857142857143e-05, |
|
"loss": 0.3518, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"learning_rate": 1.4285714285714287e-05, |
|
"loss": 0.3117, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_LOC_f1": 0.875703564727955, |
|
"eval_ORG_f1": 0.8162930563149262, |
|
"eval_PER_f1": 0.8648328896776102, |
|
"eval_loss": 0.25353190302848816, |
|
"eval_overall_accuracy": 0.9142857142857143, |
|
"eval_overall_f1": 0.8532248075732106, |
|
"eval_overall_precision": 0.8519434628975265, |
|
"eval_overall_recall": 0.8545100124047492, |
|
"eval_runtime": 7.6113, |
|
"eval_samples_per_second": 131.384, |
|
"eval_steps_per_second": 8.277, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"learning_rate": 1.3142857142857145e-05, |
|
"loss": 0.2388, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"learning_rate": 1.2e-05, |
|
"loss": 0.2468, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_LOC_f1": 0.8871085214857976, |
|
"eval_ORG_f1": 0.8523093447905478, |
|
"eval_PER_f1": 0.8841001747233547, |
|
"eval_loss": 0.21500876545906067, |
|
"eval_overall_accuracy": 0.9272277227722773, |
|
"eval_overall_f1": 0.8747007182761373, |
|
"eval_overall_precision": 0.8753993610223643, |
|
"eval_overall_recall": 0.8740031897926634, |
|
"eval_runtime": 7.4281, |
|
"eval_samples_per_second": 134.624, |
|
"eval_steps_per_second": 8.481, |
|
"step": 3750 |
|
}, |
|
{ |
|
"epoch": 3.2, |
|
"learning_rate": 1.0857142857142858e-05, |
|
"loss": 0.2071, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 3.6, |
|
"learning_rate": 9.714285714285715e-06, |
|
"loss": 0.1688, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"learning_rate": 8.571428571428571e-06, |
|
"loss": 0.1752, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_LOC_f1": 0.902119552274351, |
|
"eval_ORG_f1": 0.870843000773395, |
|
"eval_PER_f1": 0.8982737361282367, |
|
"eval_loss": 0.20536212623119354, |
|
"eval_overall_accuracy": 0.936987270155587, |
|
"eval_overall_f1": 0.890302066772655, |
|
"eval_overall_precision": 0.8874801901743264, |
|
"eval_overall_recall": 0.8931419457735247, |
|
"eval_runtime": 7.3438, |
|
"eval_samples_per_second": 136.17, |
|
"eval_steps_per_second": 8.579, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 4.4, |
|
"learning_rate": 7.428571428571429e-06, |
|
"loss": 0.1226, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 4.8, |
|
"learning_rate": 6.285714285714286e-06, |
|
"loss": 0.1241, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_LOC_f1": 0.9250525332710716, |
|
"eval_ORG_f1": 0.9045801526717557, |
|
"eval_PER_f1": 0.9203379601689801, |
|
"eval_loss": 0.16890692710876465, |
|
"eval_overall_accuracy": 0.9524045261669024, |
|
"eval_overall_f1": 0.9169995561473592, |
|
"eval_overall_precision": 0.9187122020633227, |
|
"eval_overall_recall": 0.9152932837143364, |
|
"eval_runtime": 7.4827, |
|
"eval_samples_per_second": 133.642, |
|
"eval_steps_per_second": 8.419, |
|
"step": 6250 |
|
} |
|
], |
|
"max_steps": 8750, |
|
"num_train_epochs": 7, |
|
"total_flos": 212723461717536.0, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|