|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 3.0, |
|
"eval_steps": 500, |
|
"global_step": 18, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.16666666666666666, |
|
"grad_norm": 222.38610696531117, |
|
"learning_rate": 5e-06, |
|
"loss": 2.468, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.3333333333333333, |
|
"grad_norm": 190.69596234052602, |
|
"learning_rate": 1e-05, |
|
"loss": 2.1111, |
|
"step": 2 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 44.09680989725686, |
|
"learning_rate": 9.903926402016153e-06, |
|
"loss": 1.0182, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 0.6666666666666666, |
|
"grad_norm": 14.51788378110566, |
|
"learning_rate": 9.619397662556434e-06, |
|
"loss": 0.7501, |
|
"step": 4 |
|
}, |
|
{ |
|
"epoch": 0.8333333333333334, |
|
"grad_norm": 7.591929306936709, |
|
"learning_rate": 9.157348061512728e-06, |
|
"loss": 0.4674, |
|
"step": 5 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 26.48229333375104, |
|
"learning_rate": 8.535533905932739e-06, |
|
"loss": 0.473, |
|
"step": 6 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_loss": 0.3578263819217682, |
|
"eval_runtime": 1.1002, |
|
"eval_samples_per_second": 8.18, |
|
"eval_steps_per_second": 2.727, |
|
"step": 6 |
|
}, |
|
{ |
|
"epoch": 1.1666666666666667, |
|
"grad_norm": 3.9448221983664067, |
|
"learning_rate": 7.777851165098012e-06, |
|
"loss": 0.2849, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 1.3333333333333333, |
|
"grad_norm": 3.165522164346078, |
|
"learning_rate": 6.913417161825449e-06, |
|
"loss": 0.2647, |
|
"step": 8 |
|
}, |
|
{ |
|
"epoch": 1.5, |
|
"grad_norm": 1.54216157189327, |
|
"learning_rate": 5.975451610080643e-06, |
|
"loss": 0.2519, |
|
"step": 9 |
|
}, |
|
{ |
|
"epoch": 1.6666666666666665, |
|
"grad_norm": 3.9980031503763436, |
|
"learning_rate": 5e-06, |
|
"loss": 0.2502, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 1.8333333333333335, |
|
"grad_norm": 1.88646524715919, |
|
"learning_rate": 4.02454838991936e-06, |
|
"loss": 0.2379, |
|
"step": 11 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 1.9307742582721177, |
|
"learning_rate": 3.0865828381745515e-06, |
|
"loss": 0.2043, |
|
"step": 12 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_loss": 0.27517953515052795, |
|
"eval_runtime": 0.9915, |
|
"eval_samples_per_second": 9.078, |
|
"eval_steps_per_second": 3.026, |
|
"step": 12 |
|
}, |
|
{ |
|
"epoch": 2.1666666666666665, |
|
"grad_norm": 1.4947040279267452, |
|
"learning_rate": 2.2221488349019903e-06, |
|
"loss": 0.1995, |
|
"step": 13 |
|
}, |
|
{ |
|
"epoch": 2.3333333333333335, |
|
"grad_norm": 1.0110680641273413, |
|
"learning_rate": 1.4644660940672628e-06, |
|
"loss": 0.1832, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 2.5, |
|
"grad_norm": 0.8856807961821124, |
|
"learning_rate": 8.426519384872733e-07, |
|
"loss": 0.2052, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 2.6666666666666665, |
|
"grad_norm": 1.0710608285659757, |
|
"learning_rate": 3.8060233744356634e-07, |
|
"loss": 0.2218, |
|
"step": 16 |
|
}, |
|
{ |
|
"epoch": 2.8333333333333335, |
|
"grad_norm": 0.9251900139099846, |
|
"learning_rate": 9.607359798384785e-08, |
|
"loss": 0.1858, |
|
"step": 17 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"grad_norm": 0.8991283485990265, |
|
"learning_rate": 0.0, |
|
"loss": 0.2071, |
|
"step": 18 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_loss": 0.2581753432750702, |
|
"eval_runtime": 1.0062, |
|
"eval_samples_per_second": 8.944, |
|
"eval_steps_per_second": 2.981, |
|
"step": 18 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"step": 18, |
|
"total_flos": 3943670415360.0, |
|
"train_loss": 0.554683489104112, |
|
"train_runtime": 936.1096, |
|
"train_samples_per_second": 0.542, |
|
"train_steps_per_second": 0.019 |
|
} |
|
], |
|
"logging_steps": 1, |
|
"max_steps": 18, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 3, |
|
"save_steps": 500, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": true |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 3943670415360.0, |
|
"train_batch_size": 8, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|