| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9974025974025974, | |
| "eval_steps": 6, | |
| "global_step": 24, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04155844155844156, | |
| "grad_norm": 21.0, | |
| "learning_rate": 2.5e-06, | |
| "loss": 3.3357, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04155844155844156, | |
| "eval_loss": 4.24916934967041, | |
| "eval_runtime": 139.9881, | |
| "eval_samples_per_second": 1.586, | |
| "eval_steps_per_second": 1.586, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.08311688311688312, | |
| "grad_norm": 20.625, | |
| "learning_rate": 5e-06, | |
| "loss": 3.2927, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.12467532467532468, | |
| "grad_norm": 19.125, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 3.3517, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.16623376623376623, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1e-05, | |
| "loss": 3.1756, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 11.0, | |
| "learning_rate": 9.99708511183087e-06, | |
| "loss": 3.0295, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.24935064935064935, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 9.988343845952697e-06, | |
| "loss": 2.9892, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.24935064935064935, | |
| "eval_loss": 3.628525733947754, | |
| "eval_runtime": 135.8923, | |
| "eval_samples_per_second": 1.634, | |
| "eval_steps_per_second": 1.634, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 9.973786394290475e-06, | |
| "loss": 3.0567, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.33246753246753247, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 9.953429730181653e-06, | |
| "loss": 2.9789, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.37402597402597404, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 9.927297588585984e-06, | |
| "loss": 2.9467, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 2.875, | |
| "learning_rate": 9.895420438411616e-06, | |
| "loss": 2.8889, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 9.857835446989708e-06, | |
| "loss": 2.9016, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4987012987012987, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 9.814586436738998e-06, | |
| "loss": 2.7364, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4987012987012987, | |
| "eval_loss": 3.4675309658050537, | |
| "eval_runtime": 135.783, | |
| "eval_samples_per_second": 1.635, | |
| "eval_steps_per_second": 1.635, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5402597402597402, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 9.765723834070805e-06, | |
| "loss": 2.7602, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 9.711304610594104e-06, | |
| "loss": 2.8655, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 9.651392216689167e-06, | |
| "loss": 2.7857, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6649350649350649, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 9.586056507527266e-06, | |
| "loss": 2.6186, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.7064935064935065, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 9.515373661622665e-06, | |
| "loss": 2.6869, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7480519480519481, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 9.439426092011877e-06, | |
| "loss": 2.7076, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7480519480519481, | |
| "eval_loss": 3.3928399085998535, | |
| "eval_runtime": 135.8425, | |
| "eval_samples_per_second": 1.634, | |
| "eval_steps_per_second": 1.634, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7896103896103897, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 9.358302350163758e-06, | |
| "loss": 2.6393, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 9.272097022732444e-06, | |
| "loss": 2.5202, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 2.25, | |
| "learning_rate": 9.180910621273555e-06, | |
| "loss": 2.8021, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 9.08484946505221e-06, | |
| "loss": 2.679, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.9558441558441558, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.984025557079523e-06, | |
| "loss": 2.684, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9974025974025974, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 8.8785564535221e-06, | |
| "loss": 2.757, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9974025974025974, | |
| "eval_loss": 3.348358631134033, | |
| "eval_runtime": 135.8159, | |
| "eval_samples_per_second": 1.635, | |
| "eval_steps_per_second": 1.635, | |
| "step": 24 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 96, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 24, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.773869116175155e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |