| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5111821086261981, | |
| "eval_steps": 500, | |
| "global_step": 10, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "entropy": 1.3193361535668373, | |
| "epoch": 0.051118210862619806, | |
| "grad_norm": 0.8119011521339417, | |
| "learning_rate": 0.0, | |
| "loss": 1.692, | |
| "mean_token_accuracy": 0.654717817902565, | |
| "num_tokens": 133947.0, | |
| "step": 1 | |
| }, | |
| { | |
| "entropy": 1.3207192197442055, | |
| "epoch": 0.10223642172523961, | |
| "grad_norm": 0.8001739382743835, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.6938, | |
| "mean_token_accuracy": 0.6540814265608788, | |
| "num_tokens": 267949.0, | |
| "step": 2 | |
| }, | |
| { | |
| "entropy": 1.3139144703745842, | |
| "epoch": 0.15335463258785942, | |
| "grad_norm": 0.8021382689476013, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.6875, | |
| "mean_token_accuracy": 0.6542951986193657, | |
| "num_tokens": 402435.0, | |
| "step": 3 | |
| }, | |
| { | |
| "entropy": 1.3238477781414986, | |
| "epoch": 0.20447284345047922, | |
| "grad_norm": 0.8046473264694214, | |
| "learning_rate": 6e-06, | |
| "loss": 1.6979, | |
| "mean_token_accuracy": 0.6523041352629662, | |
| "num_tokens": 536339.0, | |
| "step": 4 | |
| }, | |
| { | |
| "entropy": 1.3151762038469315, | |
| "epoch": 0.25559105431309903, | |
| "grad_norm": 0.7999162077903748, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.6884, | |
| "mean_token_accuracy": 0.6541883014142513, | |
| "num_tokens": 670396.0, | |
| "step": 5 | |
| }, | |
| { | |
| "entropy": 1.3171968907117844, | |
| "epoch": 0.30670926517571884, | |
| "grad_norm": 0.8070191740989685, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6881, | |
| "mean_token_accuracy": 0.6536356993019581, | |
| "num_tokens": 804575.0, | |
| "step": 6 | |
| }, | |
| { | |
| "entropy": 1.3177252262830734, | |
| "epoch": 0.35782747603833864, | |
| "grad_norm": 0.8115559220314026, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.6852, | |
| "mean_token_accuracy": 0.6539545804262161, | |
| "num_tokens": 938351.0, | |
| "step": 7 | |
| }, | |
| { | |
| "entropy": 1.3142458871006966, | |
| "epoch": 0.40894568690095845, | |
| "grad_norm": 0.809145987033844, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.6739, | |
| "mean_token_accuracy": 0.6544475704431534, | |
| "num_tokens": 1072655.0, | |
| "step": 8 | |
| }, | |
| { | |
| "entropy": 1.3176514655351639, | |
| "epoch": 0.46006389776357826, | |
| "grad_norm": 0.8079097270965576, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.6733, | |
| "mean_token_accuracy": 0.6541223935782909, | |
| "num_tokens": 1206472.0, | |
| "step": 9 | |
| }, | |
| { | |
| "entropy": 1.3175865784287453, | |
| "epoch": 0.5111821086261981, | |
| "grad_norm": 0.8009534478187561, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.657, | |
| "mean_token_accuracy": 0.655005007982254, | |
| "num_tokens": 1340462.0, | |
| "step": 10 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.169796942471168e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |