| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 13336, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.067293643951416, | |
| "learning_rate": 9.625074985003e-06, | |
| "loss": 1.3508, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.026179313659668, | |
| "learning_rate": 9.250149970005999e-06, | |
| "loss": 1.3055, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 6.276480674743652, | |
| "learning_rate": 8.875224955008999e-06, | |
| "loss": 1.2235, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 5.172286510467529, | |
| "learning_rate": 8.500299940011997e-06, | |
| "loss": 1.2637, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 5.915102958679199, | |
| "learning_rate": 8.125374925014997e-06, | |
| "loss": 1.2232, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.094091415405273, | |
| "learning_rate": 7.750449910017997e-06, | |
| "loss": 1.1814, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 4.9657793045043945, | |
| "learning_rate": 7.375524895020996e-06, | |
| "loss": 1.1449, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 5.007996082305908, | |
| "learning_rate": 7.000599880023996e-06, | |
| "loss": 1.0923, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 5.494340896606445, | |
| "learning_rate": 6.6256748650269955e-06, | |
| "loss": 1.0908, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 4.9410271644592285, | |
| "learning_rate": 6.250749850029995e-06, | |
| "loss": 1.0804, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 5.13407564163208, | |
| "learning_rate": 5.875824835032994e-06, | |
| "loss": 1.0556, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 4.397137641906738, | |
| "learning_rate": 5.500899820035993e-06, | |
| "loss": 1.0886, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 5.924018383026123, | |
| "learning_rate": 5.125974805038992e-06, | |
| "loss": 1.0694, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 3.952533006668091, | |
| "learning_rate": 4.751049790041992e-06, | |
| "loss": 1.0158, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 6.275745868682861, | |
| "learning_rate": 4.376124775044991e-06, | |
| "loss": 1.0082, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 7.6413116455078125, | |
| "learning_rate": 4.001199760047991e-06, | |
| "loss": 0.9991, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 5.2266387939453125, | |
| "learning_rate": 3.6262747450509898e-06, | |
| "loss": 0.976, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 4.824028968811035, | |
| "learning_rate": 3.2513497300539893e-06, | |
| "loss": 0.9826, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 7.193837642669678, | |
| "learning_rate": 2.876424715056989e-06, | |
| "loss": 0.9731, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 6.571595191955566, | |
| "learning_rate": 2.5014997000599884e-06, | |
| "loss": 1.0054, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 4.61974573135376, | |
| "learning_rate": 2.1265746850629876e-06, | |
| "loss": 0.9583, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 5.337657451629639, | |
| "learning_rate": 1.751649670065987e-06, | |
| "loss": 0.9296, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 5.575818061828613, | |
| "learning_rate": 1.3767246550689864e-06, | |
| "loss": 0.9576, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 6.3436431884765625, | |
| "learning_rate": 1.0017996400719856e-06, | |
| "loss": 0.9241, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 3.276711940765381, | |
| "learning_rate": 6.26874625074985e-07, | |
| "loss": 0.9307, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 5.67604923248291, | |
| "learning_rate": 2.519496100779844e-07, | |
| "loss": 0.9255, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 13336, | |
| "total_flos": 1.66894187000832e+16, | |
| "train_loss": 1.0640255470939504, | |
| "train_runtime": 2426.9638, | |
| "train_samples_per_second": 16.481, | |
| "train_steps_per_second": 5.495 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 13336, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 4000, | |
| "total_flos": 1.66894187000832e+16, | |
| "train_batch_size": 3, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |