| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 50.0, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 4.5681818181818186e-05, | |
| "loss": 8.5105, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_bleu": 1.0483, | |
| "eval_gen_len": 32.2222, | |
| "eval_loss": 5.73660945892334, | |
| "eval_runtime": 42.2905, | |
| "eval_samples_per_second": 2.341, | |
| "eval_steps_per_second": 0.166, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 3.431818181818182e-05, | |
| "loss": 1.3079, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_bleu": 3.8281, | |
| "eval_gen_len": 17.3838, | |
| "eval_loss": 7.149746417999268, | |
| "eval_runtime": 7.352, | |
| "eval_samples_per_second": 13.466, | |
| "eval_steps_per_second": 0.952, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 2.2954545454545457e-05, | |
| "loss": 0.179, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_bleu": 4.1437, | |
| "eval_gen_len": 18.6869, | |
| "eval_loss": 7.717140197753906, | |
| "eval_runtime": 7.7086, | |
| "eval_samples_per_second": 12.843, | |
| "eval_steps_per_second": 0.908, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 1.159090909090909e-05, | |
| "loss": 0.0535, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_bleu": 4.1251, | |
| "eval_gen_len": 18.5455, | |
| "eval_loss": 7.988112926483154, | |
| "eval_runtime": 7.5917, | |
| "eval_samples_per_second": 13.041, | |
| "eval_steps_per_second": 0.922, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 2.2727272727272726e-07, | |
| "loss": 0.0203, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_bleu": 3.9193, | |
| "eval_gen_len": 19.5758, | |
| "eval_loss": 8.036246299743652, | |
| "eval_runtime": 7.7737, | |
| "eval_samples_per_second": 12.735, | |
| "eval_steps_per_second": 0.9, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 2500, | |
| "total_flos": 8.668519071744e+16, | |
| "train_loss": 2.014233316040039, | |
| "train_runtime": 1765.0369, | |
| "train_samples_per_second": 22.634, | |
| "train_steps_per_second": 1.416 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2500, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "total_flos": 8.668519071744e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |