| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 3180, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9968553459119497, | |
| "grad_norm": 0.731029212474823, | |
| "learning_rate": 1.8006289308176103e-05, | |
| "loss": 0.408, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5906451612903226, | |
| "eval_loss": 0.20307187736034393, | |
| "eval_runtime": 1.6907, | |
| "eval_samples_per_second": 1833.602, | |
| "eval_steps_per_second": 38.446, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.4595111310482025, | |
| "learning_rate": 1.6012578616352204e-05, | |
| "loss": 0.1697, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8306451612903226, | |
| "eval_loss": 0.09759204834699631, | |
| "eval_runtime": 1.7036, | |
| "eval_samples_per_second": 1819.686, | |
| "eval_steps_per_second": 38.155, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.43162059783935547, | |
| "learning_rate": 1.4018867924528304e-05, | |
| "loss": 0.104, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.88, | |
| "eval_loss": 0.06371938437223434, | |
| "eval_runtime": 1.7221, | |
| "eval_samples_per_second": 1800.179, | |
| "eval_steps_per_second": 37.746, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.3561622202396393, | |
| "learning_rate": 1.2025157232704403e-05, | |
| "loss": 0.0773, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8977419354838709, | |
| "eval_loss": 0.04856124520301819, | |
| "eval_runtime": 1.7777, | |
| "eval_samples_per_second": 1743.857, | |
| "eval_steps_per_second": 36.565, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.2814004123210907, | |
| "learning_rate": 1.0031446540880504e-05, | |
| "loss": 0.0631, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9067741935483871, | |
| "eval_loss": 0.0395587794482708, | |
| "eval_runtime": 1.7657, | |
| "eval_samples_per_second": 1755.721, | |
| "eval_steps_per_second": 36.814, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.39958587288856506, | |
| "learning_rate": 8.037735849056606e-06, | |
| "loss": 0.0544, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9183870967741935, | |
| "eval_loss": 0.03417970985174179, | |
| "eval_runtime": 1.8, | |
| "eval_samples_per_second": 1722.246, | |
| "eval_steps_per_second": 36.112, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.27226564288139343, | |
| "learning_rate": 6.044025157232704e-06, | |
| "loss": 0.0493, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9248387096774193, | |
| "eval_loss": 0.031123189255595207, | |
| "eval_runtime": 1.7804, | |
| "eval_samples_per_second": 1741.143, | |
| "eval_steps_per_second": 36.508, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.2539806067943573, | |
| "learning_rate": 4.0503144654088055e-06, | |
| "loss": 0.0459, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9258064516129032, | |
| "eval_loss": 0.029144473373889923, | |
| "eval_runtime": 1.8058, | |
| "eval_samples_per_second": 1716.732, | |
| "eval_steps_per_second": 35.996, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.24147309362888336, | |
| "learning_rate": 2.056603773584906e-06, | |
| "loss": 0.0438, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9270967741935484, | |
| "eval_loss": 0.028016844764351845, | |
| "eval_runtime": 1.7889, | |
| "eval_samples_per_second": 1732.923, | |
| "eval_steps_per_second": 36.335, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.20841307938098907, | |
| "learning_rate": 6.289308176100629e-08, | |
| "loss": 0.0426, | |
| "step": 3170 | |
| } | |
| ], | |
| "logging_steps": 317, | |
| "max_steps": 3180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 1000000000.0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 825254092458012.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.47967279863257795, | |
| "num_train_epochs": 10, | |
| "temperature": 9 | |
| } | |
| } | |