|
{ |
|
"best_metric": 0.4463141858577728, |
|
"best_model_checkpoint": "mikhail-panzo/fil_b32_le3_s4000/checkpoint-500", |
|
"epoch": 22.22222222222222, |
|
"eval_steps": 500, |
|
"global_step": 1000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.1111111111111112, |
|
"grad_norm": 3.548590898513794, |
|
"learning_rate": 2.5e-05, |
|
"loss": 0.7354, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 2.2222222222222223, |
|
"grad_norm": 5.517350196838379, |
|
"learning_rate": 4.9500000000000004e-05, |
|
"loss": 0.5892, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 3.3333333333333335, |
|
"grad_norm": 4.143002033233643, |
|
"learning_rate": 7.45e-05, |
|
"loss": 0.532, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 4.444444444444445, |
|
"grad_norm": 6.414477825164795, |
|
"learning_rate": 9.95e-05, |
|
"loss": 0.5284, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 5.555555555555555, |
|
"grad_norm": 2.265143394470215, |
|
"learning_rate": 0.0001245, |
|
"loss": 0.517, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 6.666666666666667, |
|
"grad_norm": 4.954311847686768, |
|
"learning_rate": 0.0001495, |
|
"loss": 0.5295, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 7.777777777777778, |
|
"grad_norm": 2.5993571281433105, |
|
"learning_rate": 0.00017449999999999999, |
|
"loss": 0.517, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 8.88888888888889, |
|
"grad_norm": 3.27908992767334, |
|
"learning_rate": 0.00019950000000000002, |
|
"loss": 0.4939, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"grad_norm": 4.22811222076416, |
|
"learning_rate": 0.0002245, |
|
"loss": 0.5043, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 11.11111111111111, |
|
"grad_norm": 2.7051336765289307, |
|
"learning_rate": 0.0002495, |
|
"loss": 0.5028, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 11.11111111111111, |
|
"eval_loss": 0.4463141858577728, |
|
"eval_runtime": 16.8864, |
|
"eval_samples_per_second": 9.416, |
|
"eval_steps_per_second": 1.184, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 12.222222222222221, |
|
"grad_norm": 3.0811643600463867, |
|
"learning_rate": 0.0002745, |
|
"loss": 0.545, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 13.333333333333334, |
|
"grad_norm": 7.707492828369141, |
|
"learning_rate": 0.0002995, |
|
"loss": 0.4979, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 14.444444444444445, |
|
"grad_norm": 5.85659122467041, |
|
"learning_rate": 0.00032450000000000003, |
|
"loss": 0.4916, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 15.555555555555555, |
|
"grad_norm": 8.300439834594727, |
|
"learning_rate": 0.0003495, |
|
"loss": 0.5112, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 16.666666666666668, |
|
"grad_norm": 10.664773941040039, |
|
"learning_rate": 0.0003745, |
|
"loss": 0.5312, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 17.77777777777778, |
|
"grad_norm": 6.4097161293029785, |
|
"learning_rate": 0.0003995, |
|
"loss": 0.6087, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 18.88888888888889, |
|
"grad_norm": 5.435739994049072, |
|
"learning_rate": 0.0004245, |
|
"loss": 0.6127, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"grad_norm": 7.4649834632873535, |
|
"learning_rate": 0.00044950000000000003, |
|
"loss": 0.6904, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 21.11111111111111, |
|
"grad_norm": 10.81910228729248, |
|
"learning_rate": 0.0004745, |
|
"loss": 0.6948, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 22.22222222222222, |
|
"grad_norm": 4.4550886154174805, |
|
"learning_rate": 0.0004995, |
|
"loss": 0.6348, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 22.22222222222222, |
|
"eval_loss": 0.5990382432937622, |
|
"eval_runtime": 9.0191, |
|
"eval_samples_per_second": 17.629, |
|
"eval_steps_per_second": 2.218, |
|
"step": 1000 |
|
} |
|
], |
|
"logging_steps": 50, |
|
"max_steps": 4000, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 89, |
|
"save_steps": 500, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": false |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 6884217255754080.0, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|