| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 25, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 5.582047512699039, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.832, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.008045063224414, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.8909, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 5.619691260140896, | |
| "learning_rate": 1e-05, | |
| "loss": 0.895, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.4419954363812546, | |
| "learning_rate": 9.949107209404664e-06, | |
| "loss": 0.8295, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.388419710102024, | |
| "learning_rate": 9.797464868072489e-06, | |
| "loss": 0.8042, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 4.783176654227559, | |
| "learning_rate": 9.548159976772593e-06, | |
| "loss": 0.7827, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 5.091132431503279, | |
| "learning_rate": 9.206267664155906e-06, | |
| "loss": 0.7791, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 3.822231767141275, | |
| "learning_rate": 8.778747871771293e-06, | |
| "loss": 0.7468, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.5817851399644267, | |
| "learning_rate": 8.274303669726427e-06, | |
| "loss": 0.7373, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6946946923157924, | |
| "learning_rate": 7.703204087277989e-06, | |
| "loss": 0.6852, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.6248909167963212, | |
| "learning_rate": 7.0770750650094335e-06, | |
| "loss": 0.6784, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.5707404220045935, | |
| "learning_rate": 6.408662784207149e-06, | |
| "loss": 0.6349, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.9743312920477364, | |
| "learning_rate": 5.711574191366427e-06, | |
| "loss": 0.6179, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.1820747823709539, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6468, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0613411323142574, | |
| "learning_rate": 4.2884258086335755e-06, | |
| "loss": 0.6264, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.9343671083929717, | |
| "learning_rate": 3.5913372157928515e-06, | |
| "loss": 0.5678, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.825169270658582, | |
| "learning_rate": 2.9229249349905686e-06, | |
| "loss": 0.6467, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.7335904450059882, | |
| "learning_rate": 2.296795912722014e-06, | |
| "loss": 0.5901, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.7118131164733261, | |
| "learning_rate": 1.7256963302735752e-06, | |
| "loss": 0.5917, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5995644006563385, | |
| "learning_rate": 1.2212521282287093e-06, | |
| "loss": 0.5287, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.5329284210983056, | |
| "learning_rate": 7.937323358440935e-07, | |
| "loss": 0.5862, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.49244683959100055, | |
| "learning_rate": 4.5184002322740784e-07, | |
| "loss": 0.5381, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.5407056028850911, | |
| "learning_rate": 2.0253513192751374e-07, | |
| "loss": 0.5601, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.48949102048400017, | |
| "learning_rate": 5.089279059533658e-08, | |
| "loss": 0.5729, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5481083314386501, | |
| "learning_rate": 0.0, | |
| "loss": 0.5641, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 25, | |
| "total_flos": 3.2478016827293696e+16, | |
| "train_loss": 0.6773358273506165, | |
| "train_runtime": 2116.0034, | |
| "train_samples_per_second": 0.747, | |
| "train_steps_per_second": 0.012 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 25, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.2478016827293696e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |