| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 2060, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4854368932038835, | |
| "grad_norm": 0.5345714688301086, | |
| "learning_rate": 8.786407766990292e-05, | |
| "loss": 0.7717, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.970873786407767, | |
| "grad_norm": 0.4006052017211914, | |
| "learning_rate": 7.572815533980583e-05, | |
| "loss": 0.3916, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.4563106796116505, | |
| "grad_norm": 0.4070431888103485, | |
| "learning_rate": 6.359223300970875e-05, | |
| "loss": 0.3636, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.941747572815534, | |
| "grad_norm": 0.3988136351108551, | |
| "learning_rate": 5.145631067961165e-05, | |
| "loss": 0.3502, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.4271844660194173, | |
| "grad_norm": 0.41053929924964905, | |
| "learning_rate": 3.9320388349514564e-05, | |
| "loss": 0.3365, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.912621359223301, | |
| "grad_norm": 0.516011655330658, | |
| "learning_rate": 2.7184466019417475e-05, | |
| "loss": 0.3224, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.3980582524271843, | |
| "grad_norm": 0.5106499791145325, | |
| "learning_rate": 1.5048543689320387e-05, | |
| "loss": 0.3013, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.883495145631068, | |
| "grad_norm": 0.528724730014801, | |
| "learning_rate": 2.912621359223301e-06, | |
| "loss": 0.2887, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 2060, | |
| "total_flos": 1.4514285460762153e+18, | |
| "train_loss": 0.3876653402754404, | |
| "train_runtime": 17986.4016, | |
| "train_samples_per_second": 29.319, | |
| "train_steps_per_second": 0.115 | |
| } | |
| ], | |
| "logging_steps": 250, | |
| "max_steps": 2060, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4514285460762153e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |