| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.986666666666667, | |
| "eval_steps": 500, | |
| "global_step": 336, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.910912275314331, | |
| "learning_rate": 0.0001990856842874641, | |
| "loss": 5.0218, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.6376070380210876, | |
| "learning_rate": 0.00019297764858882514, | |
| "loss": 1.0686, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.47863319516181946, | |
| "learning_rate": 0.00018146608991420534, | |
| "loss": 0.9677, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.5812810063362122, | |
| "learning_rate": 0.0001652200182109602, | |
| "loss": 0.9866, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.47460076212882996, | |
| "learning_rate": 0.0001451835961144145, | |
| "loss": 0.8523, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.7499749660491943, | |
| "learning_rate": 0.00012252126764738844, | |
| "loss": 0.7033, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.5658572912216187, | |
| "learning_rate": 9.855008496617327e-05, | |
| "loss": 0.6454, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.7951477766036987, | |
| "learning_rate": 7.466316607649738e-05, | |
| "loss": 0.6558, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.523859441280365, | |
| "learning_rate": 5.2248731878811365e-05, | |
| "loss": 0.6391, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.6039080023765564, | |
| "learning_rate": 3.2609427815531426e-05, | |
| "loss": 0.4419, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.5373795032501221, | |
| "learning_rate": 1.6886618852849724e-05, | |
| "loss": 0.4454, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.7341424226760864, | |
| "learning_rate": 5.994057497592031e-06, | |
| "loss": 0.4096, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.7269922494888306, | |
| "learning_rate": 5.647798228764156e-07, | |
| "loss": 0.3912, | |
| "step": 325 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 336, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 1.124157764164608e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |