| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.651162790697675, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 0.0002, | |
| "loss": 4.4247, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.00017777777777777779, | |
| "loss": 2.8071, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 0.00015555555555555556, | |
| "loss": 2.2939, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 2.0343, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 1.8317, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "eval_bleu": 3.1525, | |
| "eval_gen_len": 18.6295, | |
| "eval_loss": 2.4803595542907715, | |
| "eval_runtime": 72.6261, | |
| "eval_samples_per_second": 151.461, | |
| "eval_steps_per_second": 1.184, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 1.6935, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 1.5874, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 1.5152, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 1.4649, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "learning_rate": 0.0, | |
| "loss": 1.4375, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "eval_bleu": 3.2043, | |
| "eval_gen_len": 18.62, | |
| "eval_loss": 2.523752212524414, | |
| "eval_runtime": 45.2036, | |
| "eval_samples_per_second": 243.343, | |
| "eval_steps_per_second": 1.903, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "step": 1000, | |
| "total_flos": 7.669099187195085e+16, | |
| "train_loss": 2.1090152435302736, | |
| "train_runtime": 458.607, | |
| "train_samples_per_second": 558.212, | |
| "train_steps_per_second": 2.181 | |
| } | |
| ], | |
| "max_steps": 1000, | |
| "num_train_epochs": 5, | |
| "total_flos": 7.669099187195085e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |