|
{ |
|
"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": 0.2411503791809082, |
|
"learning_rate": 0.0001, |
|
"loss": 2.0009, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.5856515373352855, |
|
"grad_norm": 0.22286373376846313, |
|
"learning_rate": 0.0001, |
|
"loss": 1.9538, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.8784773060029283, |
|
"grad_norm": 0.24079065024852753, |
|
"learning_rate": 0.0001, |
|
"loss": 1.9571, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.9985358711566618, |
|
"eval_accuracy": 0.45379487179487177, |
|
"eval_loss": 3.9512319564819336, |
|
"eval_runtime": 5.0647, |
|
"eval_samples_per_second": 98.722, |
|
"eval_steps_per_second": 12.439, |
|
"step": 341 |
|
}, |
|
{ |
|
"epoch": 0.9985358711566618, |
|
"eval_bleu": 0.28767194245703276, |
|
"eval_exact_match": 0.03, |
|
"eval_prefix_exact_match": 0.042, |
|
"step": 341 |
|
}, |
|
{ |
|
"epoch": 1.171303074670571, |
|
"grad_norm": 0.33546799421310425, |
|
"learning_rate": 0.0001, |
|
"loss": 1.9255, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 1.4641288433382138, |
|
"grad_norm": 0.3515666723251343, |
|
"learning_rate": 0.0001, |
|
"loss": 1.8684, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 1.7569546120058566, |
|
"grad_norm": 0.35733896493911743, |
|
"learning_rate": 0.0001, |
|
"loss": 1.8819, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.4482564102564103, |
|
"eval_loss": 4.112804412841797, |
|
"eval_runtime": 5.0848, |
|
"eval_samples_per_second": 98.333, |
|
"eval_steps_per_second": 12.39, |
|
"step": 683 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_bleu": 0.24521038067633488, |
|
"eval_exact_match": 0.02, |
|
"eval_prefix_exact_match": 0.044, |
|
"step": 683 |
|
}, |
|
{ |
|
"epoch": 2.049780380673499, |
|
"grad_norm": 0.3784458339214325, |
|
"learning_rate": 0.0001, |
|
"loss": 1.8613, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 2.342606149341142, |
|
"grad_norm": 0.46702849864959717, |
|
"learning_rate": 0.0001, |
|
"loss": 1.743, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 2.6354319180087846, |
|
"grad_norm": 0.5156407356262207, |
|
"learning_rate": 0.0001, |
|
"loss": 1.7653, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 2.9282576866764276, |
|
"grad_norm": 0.4755038619041443, |
|
"learning_rate": 0.0001, |
|
"loss": 1.7702, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 2.998535871156662, |
|
"eval_accuracy": 0.446051282051282, |
|
"eval_loss": 4.327652454376221, |
|
"eval_runtime": 5.1641, |
|
"eval_samples_per_second": 96.822, |
|
"eval_steps_per_second": 12.2, |
|
"step": 1024 |
|
}, |
|
{ |
|
"epoch": 2.998535871156662, |
|
"eval_bleu": 0.2598440457703149, |
|
"eval_exact_match": 0.018, |
|
"eval_prefix_exact_match": 0.034, |
|
"step": 1024 |
|
}, |
|
{ |
|
"epoch": 3.22108345534407, |
|
"grad_norm": 0.5288150906562805, |
|
"learning_rate": 0.0001, |
|
"loss": 1.6325, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 3.513909224011713, |
|
"grad_norm": 0.5784010887145996, |
|
"learning_rate": 0.0001, |
|
"loss": 1.5896, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 3.8067349926793557, |
|
"grad_norm": 0.6458429098129272, |
|
"learning_rate": 0.0001, |
|
"loss": 1.6163, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.44235897435897437, |
|
"eval_loss": 4.584905624389648, |
|
"eval_runtime": 4.9243, |
|
"eval_samples_per_second": 101.538, |
|
"eval_steps_per_second": 12.794, |
|
"step": 1366 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_bleu": 0.27967222841565076, |
|
"eval_exact_match": 0.01, |
|
"eval_prefix_exact_match": 0.038, |
|
"step": 1366 |
|
}, |
|
{ |
|
"epoch": 4.099560761346998, |
|
"grad_norm": 0.732841432094574, |
|
"learning_rate": 0.0001, |
|
"loss": 1.5441, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 4.392386530014641, |
|
"grad_norm": 0.7372816205024719, |
|
"learning_rate": 0.0001, |
|
"loss": 1.4142, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 4.685212298682284, |
|
"grad_norm": 0.8235049247741699, |
|
"learning_rate": 0.0001, |
|
"loss": 1.4198, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 4.978038067349927, |
|
"grad_norm": 0.7765743732452393, |
|
"learning_rate": 0.0001, |
|
"loss": 1.4427, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 4.998535871156662, |
|
"eval_accuracy": 0.43861538461538463, |
|
"eval_loss": 4.850330829620361, |
|
"eval_runtime": 4.9395, |
|
"eval_samples_per_second": 101.224, |
|
"eval_steps_per_second": 12.754, |
|
"step": 1707 |
|
}, |
|
{ |
|
"epoch": 4.998535871156662, |
|
"eval_bleu": 0.2170008640509292, |
|
"eval_exact_match": 0.006, |
|
"eval_prefix_exact_match": 0.032, |
|
"step": 1707 |
|
}, |
|
{ |
|
"epoch": 5.270863836017569, |
|
"grad_norm": 0.9119400978088379, |
|
"learning_rate": 0.0001, |
|
"loss": 1.2469, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 5.563689604685212, |
|
"grad_norm": 0.8966185450553894, |
|
"learning_rate": 0.0001, |
|
"loss": 1.2294, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 5.856515373352855, |
|
"grad_norm": 0.9629669189453125, |
|
"learning_rate": 0.0001, |
|
"loss": 1.2498, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.4348717948717949, |
|
"eval_loss": 5.09260892868042, |
|
"eval_runtime": 4.9572, |
|
"eval_samples_per_second": 100.863, |
|
"eval_steps_per_second": 12.709, |
|
"step": 2049 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_bleu": 0.20196479624851718, |
|
"eval_exact_match": 0.002, |
|
"eval_prefix_exact_match": 0.03, |
|
"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 |
|
} |
|
|