| { | |
| "best_metric": 0.014169026166200638, | |
| "best_model_checkpoint": "retrained_model/checkpoint-4092", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 4092, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 9.543731689453125, | |
| "learning_rate": 1.755620723362659e-05, | |
| "loss": 0.2119, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.3035629391670227, | |
| "learning_rate": 1.5112414467253179e-05, | |
| "loss": 0.0823, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_f1": 0.9882224048205971, | |
| "eval_loss": 0.028914300724864006, | |
| "eval_runtime": 88.6891, | |
| "eval_samples_per_second": 61.507, | |
| "eval_steps_per_second": 3.845, | |
| "step": 1364 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.03364085033535957, | |
| "learning_rate": 1.2668621700879767e-05, | |
| "loss": 0.0443, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.012753481045365334, | |
| "learning_rate": 1.0224828934506354e-05, | |
| "loss": 0.0273, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.009346585720777512, | |
| "learning_rate": 7.781036168132942e-06, | |
| "loss": 0.0203, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_f1": 0.9915416098226467, | |
| "eval_loss": 0.01713947393000126, | |
| "eval_runtime": 89.0176, | |
| "eval_samples_per_second": 61.28, | |
| "eval_steps_per_second": 3.831, | |
| "step": 2728 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.002515732077881694, | |
| "learning_rate": 5.337243401759531e-06, | |
| "loss": 0.0106, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 0.11586715281009674, | |
| "learning_rate": 2.8934506353861196e-06, | |
| "loss": 0.0124, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 0.0089911799877882, | |
| "learning_rate": 4.496578690127078e-07, | |
| "loss": 0.0097, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_f1": 0.9923497267759562, | |
| "eval_loss": 0.014169026166200638, | |
| "eval_runtime": 88.7641, | |
| "eval_samples_per_second": 61.455, | |
| "eval_steps_per_second": 3.842, | |
| "step": 4092 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4092, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 8140134066229824.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |