| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.847457627118644, | |
| "eval_steps": 500, | |
| "global_step": 42, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.3389830508474576, | |
| "grad_norm": 1.8727585077285767, | |
| "learning_rate": 4.827184371610511e-05, | |
| "loss": 3.1521, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6779661016949152, | |
| "grad_norm": 1.4997955560684204, | |
| "learning_rate": 4.332629679574566e-05, | |
| "loss": 2.7178, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0169491525423728, | |
| "grad_norm": 1.1865301132202148, | |
| "learning_rate": 3.5847093477938956e-05, | |
| "loss": 2.7124, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.3559322033898304, | |
| "grad_norm": 1.8665063381195068, | |
| "learning_rate": 2.686825233966061e-05, | |
| "loss": 3.0056, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.694915254237288, | |
| "grad_norm": 1.1309840679168701, | |
| "learning_rate": 1.7631120639727393e-05, | |
| "loss": 2.6855, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.0338983050847457, | |
| "grad_norm": 0.8969791531562805, | |
| "learning_rate": 9.412754953531663e-06, | |
| "loss": 2.3657, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.3728813559322033, | |
| "grad_norm": 1.163950800895691, | |
| "learning_rate": 3.3493649053890326e-06, | |
| "loss": 2.5569, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.711864406779661, | |
| "grad_norm": 2.0787620544433594, | |
| "learning_rate": 2.7922934437178695e-07, | |
| "loss": 2.9284, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.847457627118644, | |
| "step": 42, | |
| "total_flos": 285618906316800.0, | |
| "train_loss": 2.7561378251938593, | |
| "train_runtime": 901.2566, | |
| "train_samples_per_second": 0.782, | |
| "train_steps_per_second": 0.047 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 42, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "total_flos": 285618906316800.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |