| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 120.0, | |
| "eval_steps": 17748, | |
| "global_step": 177480, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.5154340267181396, | |
| "learning_rate": 9.000563443768313e-06, | |
| "loss": 0.2626, | |
| "step": 17748 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 6.236159801483154, | |
| "eval_runtime": 32.4206, | |
| "eval_samples_per_second": 299.655, | |
| "eval_steps_per_second": 6.817, | |
| "step": 17748 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.006092763505876064, | |
| "learning_rate": 8.001014198782963e-06, | |
| "loss": 0.057, | |
| "step": 35496 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_loss": 6.354074954986572, | |
| "eval_runtime": 32.2885, | |
| "eval_samples_per_second": 300.881, | |
| "eval_steps_per_second": 6.845, | |
| "step": 35496 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "grad_norm": 0.002155415015295148, | |
| "learning_rate": 7.0013522650439495e-06, | |
| "loss": 0.0428, | |
| "step": 53244 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_loss": 6.089537620544434, | |
| "eval_runtime": 30.7505, | |
| "eval_samples_per_second": 315.93, | |
| "eval_steps_per_second": 7.187, | |
| "step": 53244 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "grad_norm": 0.05033240467309952, | |
| "learning_rate": 6.001746675681768e-06, | |
| "loss": 0.0359, | |
| "step": 70992 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "eval_loss": 6.1070780754089355, | |
| "eval_runtime": 30.9274, | |
| "eval_samples_per_second": 314.123, | |
| "eval_steps_per_second": 7.146, | |
| "step": 70992 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 0.0004564746341202408, | |
| "learning_rate": 5.002253775073248e-06, | |
| "loss": 0.0312, | |
| "step": 88740 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "eval_loss": 5.999448776245117, | |
| "eval_runtime": 30.7851, | |
| "eval_samples_per_second": 315.575, | |
| "eval_steps_per_second": 7.179, | |
| "step": 88740 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "grad_norm": 0.007316610310226679, | |
| "learning_rate": 4.002704530087897e-06, | |
| "loss": 0.0287, | |
| "step": 106488 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "eval_loss": 5.854261875152588, | |
| "eval_runtime": 30.4575, | |
| "eval_samples_per_second": 318.969, | |
| "eval_steps_per_second": 7.256, | |
| "step": 106488 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "grad_norm": 0.5285100936889648, | |
| "learning_rate": 3.0032116294793783e-06, | |
| "loss": 0.0262, | |
| "step": 124236 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "eval_loss": 5.759538650512695, | |
| "eval_runtime": 30.8128, | |
| "eval_samples_per_second": 315.291, | |
| "eval_steps_per_second": 7.172, | |
| "step": 124236 | |
| }, | |
| { | |
| "epoch": 96.0, | |
| "grad_norm": 0.014891779981553555, | |
| "learning_rate": 2.003718728870859e-06, | |
| "loss": 0.0246, | |
| "step": 141984 | |
| }, | |
| { | |
| "epoch": 96.0, | |
| "eval_loss": 5.716710567474365, | |
| "eval_runtime": 30.804, | |
| "eval_samples_per_second": 315.381, | |
| "eval_steps_per_second": 7.174, | |
| "step": 141984 | |
| }, | |
| { | |
| "epoch": 108.0, | |
| "grad_norm": 0.26408958435058594, | |
| "learning_rate": 1.0041694838855084e-06, | |
| "loss": 0.0227, | |
| "step": 159732 | |
| }, | |
| { | |
| "epoch": 108.0, | |
| "eval_loss": 5.684043884277344, | |
| "eval_runtime": 32.4092, | |
| "eval_samples_per_second": 299.76, | |
| "eval_steps_per_second": 6.819, | |
| "step": 159732 | |
| }, | |
| { | |
| "epoch": 120.0, | |
| "grad_norm": 0.0010831266408786178, | |
| "learning_rate": 4.676583276988957e-09, | |
| "loss": 0.0215, | |
| "step": 177480 | |
| }, | |
| { | |
| "epoch": 120.0, | |
| "eval_loss": 5.624764919281006, | |
| "eval_runtime": 32.2837, | |
| "eval_samples_per_second": 300.926, | |
| "eval_steps_per_second": 6.846, | |
| "step": 177480 | |
| }, | |
| { | |
| "epoch": 120.0, | |
| "step": 177480, | |
| "total_flos": 2.803155588736635e+18, | |
| "train_loss": 0.05530680634294621, | |
| "train_runtime": 59846.6408, | |
| "train_samples_per_second": 118.561, | |
| "train_steps_per_second": 2.966 | |
| } | |
| ], | |
| "logging_steps": 17748, | |
| "max_steps": 177480, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 120, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.803155588736635e+18, | |
| "train_batch_size": 40, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |