|
{ |
|
"best_metric": 0.766125500202179, |
|
"best_model_checkpoint": "../finetuning_ckpts/LMWEP0WZ5IKWM.all/checkpoint-5400", |
|
"epoch": 2.0, |
|
"eval_steps": 500, |
|
"global_step": 5400, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.19, |
|
"learning_rate": 4.5370370370370374e-05, |
|
"loss": 1.1901, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"learning_rate": 4.074074074074074e-05, |
|
"loss": 0.9956, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"learning_rate": 3.611111111111111e-05, |
|
"loss": 0.9619, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"learning_rate": 3.148148148148148e-05, |
|
"loss": 0.9197, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"learning_rate": 2.6851851851851855e-05, |
|
"loss": 0.8854, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_gen_len": 8.192222222222222, |
|
"eval_loss": 0.786944568157196, |
|
"eval_rouge1": 57.6678, |
|
"eval_rouge2": 39.5459, |
|
"eval_rougeL": 56.3466, |
|
"eval_rougeLsum": 56.3932, |
|
"eval_runtime": 1027.5218, |
|
"eval_samples_per_second": 3.504, |
|
"eval_steps_per_second": 0.584, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"learning_rate": 2.2222222222222223e-05, |
|
"loss": 0.7933, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 1.3, |
|
"learning_rate": 1.7592592592592595e-05, |
|
"loss": 0.7255, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 1.48, |
|
"learning_rate": 1.2962962962962962e-05, |
|
"loss": 0.7384, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 1.67, |
|
"learning_rate": 8.333333333333334e-06, |
|
"loss": 0.7215, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 1.85, |
|
"learning_rate": 3.7037037037037037e-06, |
|
"loss": 0.7082, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_gen_len": 8.563055555555556, |
|
"eval_loss": 0.766125500202179, |
|
"eval_rouge1": 58.2139, |
|
"eval_rouge2": 39.2923, |
|
"eval_rougeL": 56.887, |
|
"eval_rougeLsum": 56.9112, |
|
"eval_runtime": 1041.4148, |
|
"eval_samples_per_second": 3.457, |
|
"eval_steps_per_second": 0.576, |
|
"step": 5400 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 5400, |
|
"num_train_epochs": 2, |
|
"save_steps": 500, |
|
"total_flos": 4.39203448083456e+17, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|