| { | |
| "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": 0.07, | |
| "grad_norm": 8.27767250823638, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.7254, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 354.28943373575305, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3589, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 44.47148792434896, | |
| "learning_rate": 1.911228490388136e-05, | |
| "loss": 2.2101, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "eval_loss": 1.495993971824646, | |
| "eval_runtime": 16.1012, | |
| "eval_samples_per_second": 11.552, | |
| "eval_steps_per_second": 0.745, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 7.3069076279787994, | |
| "learning_rate": 1.6606747233900816e-05, | |
| "loss": 1.0724, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 4.8728357716552235, | |
| "learning_rate": 1.2928227712765504e-05, | |
| "loss": 0.9386, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 3.6690847322663, | |
| "learning_rate": 8.729821802531213e-06, | |
| "loss": 0.8912, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "eval_loss": 1.3901439905166626, | |
| "eval_runtime": 15.9261, | |
| "eval_samples_per_second": 11.679, | |
| "eval_steps_per_second": 0.753, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 3.0703999031199913, | |
| "learning_rate": 4.756927164427685e-06, | |
| "loss": 0.8339, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 2.5427246992810155, | |
| "learning_rate": 1.7149035075615795e-06, | |
| "loss": 0.7913, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 2.549641028108478, | |
| "learning_rate": 1.4384089652291544e-07, | |
| "loss": 0.8119, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "eval_loss": 1.3799355030059814, | |
| "eval_runtime": 15.8926, | |
| "eval_samples_per_second": 11.704, | |
| "eval_steps_per_second": 0.755, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "step": 42, | |
| "total_flos": 8741600624640.0, | |
| "train_loss": 1.0784317638192857, | |
| "train_runtime": 762.3306, | |
| "train_samples_per_second": 3.715, | |
| "train_steps_per_second": 0.055 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 42, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 8741600624640.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |