| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 4460, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.8142107129096985, | |
| "learning_rate": 0.0008, | |
| "loss": 3.3525, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 2.4645345211029053, | |
| "eval_rouge1": 0.6739, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6721, | |
| "eval_rougeLsum": 0.6722, | |
| "eval_runtime": 52.0826, | |
| "eval_samples_per_second": 14.343, | |
| "eval_steps_per_second": 0.461, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7527865767478943, | |
| "learning_rate": 0.0006, | |
| "loss": 2.9432, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 2.228515625, | |
| "eval_rouge1": 0.6712, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6719, | |
| "eval_rougeLsum": 0.672, | |
| "eval_runtime": 54.6492, | |
| "eval_samples_per_second": 13.669, | |
| "eval_steps_per_second": 0.439, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.7620196342468262, | |
| "learning_rate": 0.0004, | |
| "loss": 2.7301, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 2.000340223312378, | |
| "eval_rouge1": 0.6833, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6773, | |
| "eval_rougeLsum": 0.6791, | |
| "eval_runtime": 49.9831, | |
| "eval_samples_per_second": 14.945, | |
| "eval_steps_per_second": 0.48, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.8655580878257751, | |
| "learning_rate": 0.0002, | |
| "loss": 2.5599, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.8546103239059448, | |
| "eval_rouge1": 0.7278, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7227, | |
| "eval_rougeLsum": 0.7246, | |
| "eval_runtime": 50.7437, | |
| "eval_samples_per_second": 14.721, | |
| "eval_steps_per_second": 0.473, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.7443877458572388, | |
| "learning_rate": 0.0, | |
| "loss": 2.422, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.7829684019088745, | |
| "eval_rouge1": 0.6862, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6838, | |
| "eval_rougeLsum": 0.6843, | |
| "eval_runtime": 52.337, | |
| "eval_samples_per_second": 14.273, | |
| "eval_steps_per_second": 0.459, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 5.531847691862016e+16, | |
| "train_loss": 2.801529929669983, | |
| "train_runtime": 5646.3824, | |
| "train_samples_per_second": 12.633, | |
| "train_steps_per_second": 0.79 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.531847691862016e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |