|
{ |
|
"best_metric": 0.9185779816513762, |
|
"best_model_checkpoint": "electra-distilled-sst\\run-8\\checkpoint-3162", |
|
"epoch": 6.0, |
|
"eval_steps": 500, |
|
"global_step": 3162, |
|
"is_hyper_param_search": true, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 25.336427688598633, |
|
"learning_rate": 8.271664431386432e-05, |
|
"loss": 0.94, |
|
"step": 527 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.9036697247706422, |
|
"eval_loss": 0.6703569889068604, |
|
"eval_runtime": 1.4486, |
|
"eval_samples_per_second": 601.968, |
|
"eval_steps_per_second": 4.832, |
|
"step": 527 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 11.467299461364746, |
|
"learning_rate": 7.352784339619853e-05, |
|
"loss": 0.4737, |
|
"step": 1054 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.8990825688073395, |
|
"eval_loss": 0.8164154887199402, |
|
"eval_runtime": 1.4496, |
|
"eval_samples_per_second": 601.528, |
|
"eval_steps_per_second": 4.829, |
|
"step": 1054 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"grad_norm": 44.32056427001953, |
|
"learning_rate": 6.433904247853274e-05, |
|
"loss": 0.3435, |
|
"step": 1581 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.8979357798165137, |
|
"eval_loss": 0.8911434412002563, |
|
"eval_runtime": 1.4479, |
|
"eval_samples_per_second": 602.254, |
|
"eval_steps_per_second": 4.835, |
|
"step": 1581 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"grad_norm": 6.458914756774902, |
|
"learning_rate": 5.515024156086696e-05, |
|
"loss": 0.2599, |
|
"step": 2108 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.9151376146788991, |
|
"eval_loss": 0.7757946252822876, |
|
"eval_runtime": 1.446, |
|
"eval_samples_per_second": 603.064, |
|
"eval_steps_per_second": 4.841, |
|
"step": 2108 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"grad_norm": Infinity, |
|
"learning_rate": 4.5978876698073397e-05, |
|
"loss": 0.2156, |
|
"step": 2635 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.9162844036697247, |
|
"eval_loss": 0.7059925198554993, |
|
"eval_runtime": 1.4472, |
|
"eval_samples_per_second": 602.529, |
|
"eval_steps_per_second": 4.837, |
|
"step": 2635 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"grad_norm": 26.35287857055664, |
|
"learning_rate": 3.679007578040761e-05, |
|
"loss": 0.1789, |
|
"step": 3162 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.9185779816513762, |
|
"eval_loss": 0.6810566186904907, |
|
"eval_runtime": 1.4505, |
|
"eval_samples_per_second": 601.155, |
|
"eval_steps_per_second": 4.826, |
|
"step": 3162 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 5270, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 10, |
|
"save_steps": 500, |
|
"total_flos": 1123721737396596.0, |
|
"train_batch_size": 128, |
|
"trial_name": null, |
|
"trial_params": { |
|
"alpha": 0.7393993567935025, |
|
"learning_rate": 9.188800917665787e-05, |
|
"num_train_epochs": 10, |
|
"temperature": 6 |
|
} |
|
} |
|
|