| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 14.4, | |
| "eval_steps": 500, | |
| "global_step": 180, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.8724178671836853, | |
| "learning_rate": 0.00019868265225415265, | |
| "loss": 4.9162, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.0800001621246338, | |
| "learning_rate": 0.00019075754196709572, | |
| "loss": 3.2245, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.6155961751937866, | |
| "learning_rate": 0.00017621620551276366, | |
| "loss": 1.7826, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.4888613522052765, | |
| "learning_rate": 0.00015611870653623825, | |
| "loss": 1.3024, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.41620710492134094, | |
| "learning_rate": 0.000131930153013598, | |
| "loss": 1.053, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.41758859157562256, | |
| "learning_rate": 0.00010541389085854176, | |
| "loss": 0.9255, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.3929082751274109, | |
| "learning_rate": 7.85029559788976e-05, | |
| "loss": 0.8443, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.4028986692428589, | |
| "learning_rate": 5.3159155930021e-05, | |
| "loss": 0.7949, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.3700142204761505, | |
| "learning_rate": 3.123005411465766e-05, | |
| "loss": 0.7564, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.38459262251853943, | |
| "learning_rate": 1.4314282383241096e-05, | |
| "loss": 0.7354, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "grad_norm": 0.38667017221450806, | |
| "learning_rate": 3.6450007480777093e-06, | |
| "loss": 0.7324, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "grad_norm": 0.3338981568813324, | |
| "learning_rate": 0.0, | |
| "loss": 0.7366, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "step": 180, | |
| "total_flos": 116292530995200.0, | |
| "train_loss": 1.4836891492207844, | |
| "train_runtime": 82.2898, | |
| "train_samples_per_second": 18.228, | |
| "train_steps_per_second": 2.187 | |
| } | |
| ], | |
| "logging_steps": 15, | |
| "max_steps": 180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 500, | |
| "total_flos": 116292530995200.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |