| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 4470, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.76835036277771, | |
| "learning_rate": 0.0008, | |
| "loss": 3.0441, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.9890629053115845, | |
| "eval_rouge1": 0.6966, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.698, | |
| "eval_rougeLsum": 0.6962, | |
| "eval_runtime": 53.7394, | |
| "eval_samples_per_second": 13.9, | |
| "eval_steps_per_second": 0.447, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6674537658691406, | |
| "learning_rate": 0.0006, | |
| "loss": 2.4037, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.6701865196228027, | |
| "eval_rouge1": 0.7078, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7135, | |
| "eval_rougeLsum": 0.7146, | |
| "eval_runtime": 50.3382, | |
| "eval_samples_per_second": 14.84, | |
| "eval_steps_per_second": 0.477, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.4838422536849976, | |
| "learning_rate": 0.0004, | |
| "loss": 2.1345, | |
| "step": 2682 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.4640010595321655, | |
| "eval_rouge1": 0.6592, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.66, | |
| "eval_rougeLsum": 0.6572, | |
| "eval_runtime": 51.6537, | |
| "eval_samples_per_second": 14.462, | |
| "eval_steps_per_second": 0.465, | |
| "step": 2682 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.5656119585037231, | |
| "learning_rate": 0.0002, | |
| "loss": 1.9436, | |
| "step": 3576 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.3520551919937134, | |
| "eval_rouge1": 0.6535, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6545, | |
| "eval_rougeLsum": 0.6547, | |
| "eval_runtime": 54.7923, | |
| "eval_samples_per_second": 13.633, | |
| "eval_steps_per_second": 0.438, | |
| "step": 3576 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.9242256879806519, | |
| "learning_rate": 0.0, | |
| "loss": 1.7989, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.268640398979187, | |
| "eval_rouge1": 0.6818, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6864, | |
| "eval_rougeLsum": 0.6834, | |
| "eval_runtime": 51.2431, | |
| "eval_samples_per_second": 14.578, | |
| "eval_steps_per_second": 0.468, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4470, | |
| "total_flos": 5.54115403874304e+16, | |
| "train_loss": 2.264926113324944, | |
| "train_runtime": 5745.1129, | |
| "train_samples_per_second": 12.437, | |
| "train_steps_per_second": 0.778 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4470, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.54115403874304e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |