| { | |
| "best_metric": 0.8761239051818848, | |
| "best_model_checkpoint": "./tst-translation-output/checkpoint-3000", | |
| "epoch": 1.8217701533323214, | |
| "eval_steps": 1500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 1.562, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 4.956224829276834e-05, | |
| "loss": 1.118, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 4.912449658553668e-05, | |
| "loss": 1.014, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "eval_bleu": 5.9697, | |
| "eval_gen_len": 18.2307, | |
| "eval_loss": 0.9561266303062439, | |
| "eval_runtime": 1131.8043, | |
| "eval_samples_per_second": 11.638, | |
| "eval_steps_per_second": 1.455, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 4.868674487830503e-05, | |
| "loss": 0.8557, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 4.824899317107337e-05, | |
| "loss": 0.7997, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 4.781124146384171e-05, | |
| "loss": 0.7888, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "eval_bleu": 10.9388, | |
| "eval_gen_len": 17.4752, | |
| "eval_loss": 0.8761239051818848, | |
| "eval_runtime": 1004.0048, | |
| "eval_samples_per_second": 13.119, | |
| "eval_steps_per_second": 1.64, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 57610, | |
| "num_train_epochs": 35, | |
| "save_steps": 1500, | |
| "total_flos": 4.16088915443712e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |