| { | |
| "best_metric": 1.7036641836166382, | |
| "best_model_checkpoint": "flan-t5-large/checkpoint-604", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 1208, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 106.27584586466165, | |
| "eval_loss": 1.7234793901443481, | |
| "eval_rouge1": 44.3127, | |
| "eval_rouge2": 18.5321, | |
| "eval_rougeL": 31.8618, | |
| "eval_rougeLsum": 41.1286, | |
| "eval_runtime": 1981.3194, | |
| "eval_samples_per_second": 1.064, | |
| "eval_steps_per_second": 0.019, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.43800386174930117, | |
| "learning_rate": 0.0001, | |
| "loss": 1.8444, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 98.73778195488721, | |
| "eval_loss": 1.7036641836166382, | |
| "eval_rouge1": 44.6191, | |
| "eval_rouge2": 18.529, | |
| "eval_rougeL": 32.4108, | |
| "eval_rougeLsum": 41.5764, | |
| "eval_runtime": 1977.8502, | |
| "eval_samples_per_second": 1.066, | |
| "eval_steps_per_second": 0.019, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 100.92763157894737, | |
| "eval_loss": 1.7065348625183105, | |
| "eval_rouge1": 45.2044, | |
| "eval_rouge2": 19.0224, | |
| "eval_rougeL": 32.7758, | |
| "eval_rougeLsum": 42.1475, | |
| "eval_runtime": 1977.3953, | |
| "eval_samples_per_second": 1.066, | |
| "eval_steps_per_second": 0.019, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 0.4764731580095982, | |
| "learning_rate": 0.0001, | |
| "loss": 1.6046, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 97.4515977443609, | |
| "eval_loss": 1.7114441394805908, | |
| "eval_rouge1": 45.5049, | |
| "eval_rouge2": 19.1992, | |
| "eval_rougeL": 32.905, | |
| "eval_rougeLsum": 42.5556, | |
| "eval_runtime": 1982.5948, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 0.019, | |
| "step": 1208 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1510, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 25961869344768.0, | |
| "train_batch_size": 14, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |