| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.457657657657657, | |
| "global_step": 14336, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9157, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.647285753651144e-05, | |
| "loss": 0.4765, | |
| "step": 2048 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu_score": 1.0824, | |
| "eval_loss": 0.42013075947761536, | |
| "eval_runtime": 371.3275, | |
| "eval_samples_per_second": 11.957, | |
| "eval_steps_per_second": 1.495, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 4.294571507302287e-05, | |
| "loss": 0.389, | |
| "step": 3072 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 3.941857260953431e-05, | |
| "loss": 0.3504, | |
| "step": 4096 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu_score": 3.3723, | |
| "eval_loss": 0.3275691866874695, | |
| "eval_runtime": 367.6436, | |
| "eval_samples_per_second": 12.077, | |
| "eval_steps_per_second": 1.51, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 3.589143014604575e-05, | |
| "loss": 0.2749, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 3.236428768255718e-05, | |
| "loss": 0.2359, | |
| "step": 6144 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu_score": 8.4799, | |
| "eval_loss": 0.26540276408195496, | |
| "eval_runtime": 394.9116, | |
| "eval_samples_per_second": 11.243, | |
| "eval_steps_per_second": 1.405, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 2.8837145219068618e-05, | |
| "loss": 0.1857, | |
| "step": 7168 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "learning_rate": 2.5310002755580052e-05, | |
| "loss": 0.1487, | |
| "step": 8192 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu_score": 19.2993, | |
| "eval_loss": 0.2209530770778656, | |
| "eval_runtime": 389.9194, | |
| "eval_samples_per_second": 11.387, | |
| "eval_steps_per_second": 1.423, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "learning_rate": 2.1782860292091487e-05, | |
| "loss": 0.1258, | |
| "step": 9216 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 1.8255717828602922e-05, | |
| "loss": 0.0953, | |
| "step": 10240 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu_score": 35.2489, | |
| "eval_loss": 0.19596563279628754, | |
| "eval_runtime": 396.3886, | |
| "eval_samples_per_second": 11.201, | |
| "eval_steps_per_second": 1.4, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "learning_rate": 1.4728575365114358e-05, | |
| "loss": 0.0849, | |
| "step": 11264 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "learning_rate": 1.1201432901625793e-05, | |
| "loss": 0.0605, | |
| "step": 12288 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 7.67429043813723e-06, | |
| "loss": 0.056, | |
| "step": 13312 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu_score": 52.7747, | |
| "eval_loss": 0.17908145487308502, | |
| "eval_runtime": 403.663, | |
| "eval_samples_per_second": 10.999, | |
| "eval_steps_per_second": 1.375, | |
| "step": 13320 | |
| }, | |
| { | |
| "epoch": 6.46, | |
| "learning_rate": 4.147147974648664e-06, | |
| "loss": 0.0415, | |
| "step": 14336 | |
| } | |
| ], | |
| "max_steps": 15540, | |
| "num_train_epochs": 7, | |
| "total_flos": 2.06948771780918e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |