| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 28, | |
| "global_step": 56, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 3e-05, | |
| "loss": 3.5167, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 3e-05, | |
| "loss": 2.7034, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3e-05, | |
| "loss": 2.5409, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4931, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_accuracy": 0.5748751060189339, | |
| "eval_loss": 2.390355348587036, | |
| "eval_runtime": 25.7729, | |
| "eval_samples_per_second": 34.261, | |
| "eval_steps_per_second": 2.173, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_bleu": 2.6290615444017718, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.0, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4944, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4832, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4336, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3e-05, | |
| "loss": 2.5101, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4699, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6049526162776446, | |
| "eval_loss": 2.101423740386963, | |
| "eval_runtime": 23.9781, | |
| "eval_samples_per_second": 36.825, | |
| "eval_steps_per_second": 2.335, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 2.731416805620766, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.0, | |
| "step": 56 | |
| } | |
| ], | |
| "logging_steps": 6, | |
| "max_steps": 560, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1360515013171200.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |