| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 40000, | |
| "global_step": 5400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 6.538878440856934, | |
| "learning_rate": 2.3169601482854498e-05, | |
| "loss": 1.6323, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 5.898880958557129, | |
| "learning_rate": 2.1316033364226137e-05, | |
| "loss": 0.8545, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 3.7976324558258057, | |
| "learning_rate": 1.946246524559778e-05, | |
| "loss": 0.782, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 5.201897621154785, | |
| "learning_rate": 1.7608897126969417e-05, | |
| "loss": 0.7559, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 6.1030988693237305, | |
| "learning_rate": 1.5755329008341056e-05, | |
| "loss": 0.7381, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 8.243881225585938, | |
| "learning_rate": 1.3901760889712698e-05, | |
| "loss": 0.7118, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 6.349460601806641, | |
| "learning_rate": 1.2048192771084338e-05, | |
| "loss": 0.6672, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 7.698678970336914, | |
| "learning_rate": 1.0194624652455977e-05, | |
| "loss": 0.5414, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 8.411925315856934, | |
| "learning_rate": 8.341056533827618e-06, | |
| "loss": 0.5194, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 8.214927673339844, | |
| "learning_rate": 6.487488415199258e-06, | |
| "loss": 0.5199, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 10.487695693969727, | |
| "learning_rate": 4.633920296570899e-06, | |
| "loss": 0.4974, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 12.671321868896484, | |
| "learning_rate": 2.7803521779425395e-06, | |
| "loss": 0.4919, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 6.23270320892334, | |
| "learning_rate": 9.267840593141798e-07, | |
| "loss": 0.4911, | |
| "step": 5200 | |
| } | |
| ], | |
| "logging_steps": 400, | |
| "max_steps": 5400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "total_flos": 2.0661622822760448e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |