| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.35555555555555557, | |
| "eval_steps": 10, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 3.639567466962035e-06, | |
| "learning_rate": 4.988888888888889e-06, | |
| "loss": 0.6992, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "eval_accuracy": 0.5314685314685315, | |
| "eval_loss": 0.6923828125, | |
| "eval_runtime": 17.5567, | |
| "eval_samples_per_second": 11.392, | |
| "eval_steps_per_second": 1.424, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 2.6224586963653564, | |
| "learning_rate": 4.977777777777778e-06, | |
| "loss": 0.6876, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "eval_accuracy": 0.5503355704697986, | |
| "eval_loss": 0.69140625, | |
| "eval_runtime": 17.5256, | |
| "eval_samples_per_second": 11.412, | |
| "eval_steps_per_second": 1.426, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 3.8953466415405273, | |
| "learning_rate": 4.966666666666667e-06, | |
| "loss": 0.7041, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "eval_accuracy": 0.6625, | |
| "eval_loss": 0.68505859375, | |
| "eval_runtime": 17.5123, | |
| "eval_samples_per_second": 11.421, | |
| "eval_steps_per_second": 1.428, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 2.924593448638916, | |
| "learning_rate": 4.9555555555555565e-06, | |
| "loss": 0.6805, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "eval_accuracy": 0.6081081081081081, | |
| "eval_loss": 0.689453125, | |
| "eval_runtime": 17.5121, | |
| "eval_samples_per_second": 11.421, | |
| "eval_steps_per_second": 1.428, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 6.549908903252799e-06, | |
| "learning_rate": 4.944444444444445e-06, | |
| "loss": 0.6947, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "eval_accuracy": 0.5197368421052632, | |
| "eval_loss": 0.69091796875, | |
| "eval_runtime": 17.5075, | |
| "eval_samples_per_second": 11.424, | |
| "eval_steps_per_second": 1.428, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 4.414588602230651e-06, | |
| "learning_rate": 4.933333333333334e-06, | |
| "loss": 0.6961, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "eval_accuracy": 0.5632911392405063, | |
| "eval_loss": 0.69189453125, | |
| "eval_runtime": 17.5091, | |
| "eval_samples_per_second": 11.423, | |
| "eval_steps_per_second": 1.428, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 4.562589765555458e-06, | |
| "learning_rate": 4.922222222222223e-06, | |
| "loss": 0.6813, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "eval_accuracy": 0.6282051282051282, | |
| "eval_loss": 0.68798828125, | |
| "eval_runtime": 17.51, | |
| "eval_samples_per_second": 11.422, | |
| "eval_steps_per_second": 1.428, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 2.7170662879943848, | |
| "learning_rate": 4.911111111111112e-06, | |
| "loss": 0.689, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "eval_accuracy": 0.525, | |
| "eval_loss": 0.689453125, | |
| "eval_runtime": 17.5037, | |
| "eval_samples_per_second": 11.426, | |
| "eval_steps_per_second": 1.428, | |
| "step": 80 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 4500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |