|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 6.0, |
|
"eval_steps": 500, |
|
"global_step": 2049, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.29282576866764276, |
|
"grad_norm": 0.26115021109580994, |
|
"learning_rate": 3e-05, |
|
"loss": 2.0719, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.5856515373352855, |
|
"grad_norm": 0.2679530680179596, |
|
"learning_rate": 3e-05, |
|
"loss": 1.969, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.8784773060029283, |
|
"grad_norm": 0.2916224002838135, |
|
"learning_rate": 3e-05, |
|
"loss": 1.9686, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.9985358711566618, |
|
"eval_accuracy": 0.43743589743589745, |
|
"eval_loss": 3.6918885707855225, |
|
"eval_runtime": 5.2974, |
|
"eval_samples_per_second": 94.385, |
|
"eval_steps_per_second": 11.893, |
|
"step": 341 |
|
}, |
|
{ |
|
"epoch": 0.9985358711566618, |
|
"eval_bleu": 0.1527908371131025, |
|
"eval_exact_match": 0.012, |
|
"eval_prefix_exact_match": 0.026, |
|
"step": 341 |
|
}, |
|
{ |
|
"epoch": 1.171303074670571, |
|
"grad_norm": 0.3534347712993622, |
|
"learning_rate": 3e-05, |
|
"loss": 1.9603, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 1.4641288433382138, |
|
"grad_norm": 0.3338424563407898, |
|
"learning_rate": 3e-05, |
|
"loss": 1.9235, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 1.7569546120058566, |
|
"grad_norm": 0.4033390283584595, |
|
"learning_rate": 3e-05, |
|
"loss": 1.9337, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.44764102564102565, |
|
"eval_loss": 3.7484843730926514, |
|
"eval_runtime": 5.0649, |
|
"eval_samples_per_second": 98.718, |
|
"eval_steps_per_second": 12.438, |
|
"step": 683 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_bleu": 0.21311912714932216, |
|
"eval_exact_match": 0.02, |
|
"eval_prefix_exact_match": 0.032, |
|
"step": 683 |
|
}, |
|
{ |
|
"epoch": 2.049780380673499, |
|
"grad_norm": 0.38882532715797424, |
|
"learning_rate": 3e-05, |
|
"loss": 1.9267, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 2.342606149341142, |
|
"grad_norm": 0.44310709834098816, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8823, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 2.6354319180087846, |
|
"grad_norm": 0.5196680426597595, |
|
"learning_rate": 3e-05, |
|
"loss": 1.901, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 2.9282576866764276, |
|
"grad_norm": 0.5293437242507935, |
|
"learning_rate": 3e-05, |
|
"loss": 1.9033, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 2.998535871156662, |
|
"eval_accuracy": 0.4495897435897436, |
|
"eval_loss": 3.882551670074463, |
|
"eval_runtime": 5.1314, |
|
"eval_samples_per_second": 97.439, |
|
"eval_steps_per_second": 12.277, |
|
"step": 1024 |
|
}, |
|
{ |
|
"epoch": 2.998535871156662, |
|
"eval_bleu": 0.24307488023362148, |
|
"eval_exact_match": 0.016, |
|
"eval_prefix_exact_match": 0.032, |
|
"step": 1024 |
|
}, |
|
{ |
|
"epoch": 3.22108345534407, |
|
"grad_norm": 0.552943229675293, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8529, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 3.513909224011713, |
|
"grad_norm": 0.6421984434127808, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8342, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 3.8067349926793557, |
|
"grad_norm": 0.7155198454856873, |
|
"learning_rate": 3e-05, |
|
"loss": 1.857, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.4481025641025641, |
|
"eval_loss": 3.970107316970825, |
|
"eval_runtime": 5.1563, |
|
"eval_samples_per_second": 96.969, |
|
"eval_steps_per_second": 12.218, |
|
"step": 1366 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_bleu": 0.25309406645138827, |
|
"eval_exact_match": 0.016, |
|
"eval_prefix_exact_match": 0.044, |
|
"step": 1366 |
|
}, |
|
{ |
|
"epoch": 4.099560761346998, |
|
"grad_norm": 0.7837833166122437, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8284, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 4.392386530014641, |
|
"grad_norm": 0.784609854221344, |
|
"learning_rate": 3e-05, |
|
"loss": 1.7884, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 4.685212298682284, |
|
"grad_norm": 0.834195613861084, |
|
"learning_rate": 3e-05, |
|
"loss": 1.7841, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 4.978038067349927, |
|
"grad_norm": 0.8902062177658081, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8042, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 4.998535871156662, |
|
"eval_accuracy": 0.4472820512820513, |
|
"eval_loss": 4.117065906524658, |
|
"eval_runtime": 5.2376, |
|
"eval_samples_per_second": 95.464, |
|
"eval_steps_per_second": 12.028, |
|
"step": 1707 |
|
}, |
|
{ |
|
"epoch": 4.998535871156662, |
|
"eval_bleu": 0.2531049955279983, |
|
"eval_exact_match": 0.014, |
|
"eval_prefix_exact_match": 0.042, |
|
"step": 1707 |
|
}, |
|
{ |
|
"epoch": 5.270863836017569, |
|
"grad_norm": 0.9495383501052856, |
|
"learning_rate": 3e-05, |
|
"loss": 1.7493, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 5.563689604685212, |
|
"grad_norm": 0.9647482633590698, |
|
"learning_rate": 3e-05, |
|
"loss": 1.7334, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 5.856515373352855, |
|
"grad_norm": 0.9825767874717712, |
|
"learning_rate": 3e-05, |
|
"loss": 1.7443, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.446974358974359, |
|
"eval_loss": 4.183700084686279, |
|
"eval_runtime": 5.2674, |
|
"eval_samples_per_second": 94.923, |
|
"eval_steps_per_second": 11.96, |
|
"step": 2049 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_bleu": 0.23587359955369594, |
|
"eval_exact_match": 0.012, |
|
"eval_prefix_exact_match": 0.04, |
|
"step": 2049 |
|
} |
|
], |
|
"logging_steps": 100, |
|
"max_steps": 17050, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 50, |
|
"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": 1.7594902430036787e+17, |
|
"train_batch_size": 1, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|