| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "eval_steps": 500, | |
| "global_step": 4770, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9968553459119497, | |
| "grad_norm": 0.3008987009525299, | |
| "learning_rate": 0.00021943281418587574, | |
| "loss": 0.1836, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8977419354838709, | |
| "eval_loss": 0.048293426632881165, | |
| "eval_runtime": 5.4409, | |
| "eval_samples_per_second": 569.756, | |
| "eval_steps_per_second": 11.946, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.1929541975259781, | |
| "learning_rate": 0.000212311269269539, | |
| "loss": 0.0436, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9235483870967742, | |
| "eval_loss": 0.03785553574562073, | |
| "eval_runtime": 5.4055, | |
| "eval_samples_per_second": 573.489, | |
| "eval_steps_per_second": 12.025, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.1676182746887207, | |
| "learning_rate": 0.00020078619740999677, | |
| "loss": 0.0318, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9332258064516129, | |
| "eval_loss": 0.029865220189094543, | |
| "eval_runtime": 5.3842, | |
| "eval_samples_per_second": 575.754, | |
| "eval_steps_per_second": 12.072, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.14221236109733582, | |
| "learning_rate": 0.00018535814809928703, | |
| "loss": 0.0243, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9403225806451613, | |
| "eval_loss": 0.027224214747548103, | |
| "eval_runtime": 5.4527, | |
| "eval_samples_per_second": 568.529, | |
| "eval_steps_per_second": 11.921, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.08744881302118301, | |
| "learning_rate": 0.00016669718243028002, | |
| "loss": 0.0211, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9432258064516129, | |
| "eval_loss": 0.023913592100143433, | |
| "eval_runtime": 5.4569, | |
| "eval_samples_per_second": 568.09, | |
| "eval_steps_per_second": 11.912, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.08495471626520157, | |
| "learning_rate": 0.00014561377143520847, | |
| "loss": 0.019, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9419354838709677, | |
| "eval_loss": 0.023042848333716393, | |
| "eval_runtime": 5.4211, | |
| "eval_samples_per_second": 571.836, | |
| "eval_steps_per_second": 11.99, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.05824749171733856, | |
| "learning_rate": 0.00012302359623021862, | |
| "loss": 0.0177, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.94, | |
| "eval_loss": 0.022627023980021477, | |
| "eval_runtime": 5.3959, | |
| "eval_samples_per_second": 574.514, | |
| "eval_steps_per_second": 12.046, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.07080487161874771, | |
| "learning_rate": 9.990777874337415e-05, | |
| "loss": 0.0165, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9406451612903226, | |
| "eval_loss": 0.02114924229681492, | |
| "eval_runtime": 5.4349, | |
| "eval_samples_per_second": 570.39, | |
| "eval_steps_per_second": 11.96, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.06504567712545395, | |
| "learning_rate": 7.727027025949969e-05, | |
| "loss": 0.0154, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9458064516129032, | |
| "eval_loss": 0.019661111757159233, | |
| "eval_runtime": 5.4444, | |
| "eval_samples_per_second": 569.397, | |
| "eval_steps_per_second": 11.939, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.05454040318727493, | |
| "learning_rate": 5.6094248455116685e-05, | |
| "loss": 0.0146, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9483870967741935, | |
| "eval_loss": 0.01887706108391285, | |
| "eval_runtime": 5.4106, | |
| "eval_samples_per_second": 572.947, | |
| "eval_steps_per_second": 12.013, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 10.965408805031446, | |
| "grad_norm": 0.06293663382530212, | |
| "learning_rate": 3.7299416659345335e-05, | |
| "loss": 0.0137, | |
| "step": 3487 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9487096774193549, | |
| "eval_loss": 0.0184471495449543, | |
| "eval_runtime": 5.408, | |
| "eval_samples_per_second": 573.222, | |
| "eval_steps_per_second": 12.019, | |
| "step": 3498 | |
| }, | |
| { | |
| "epoch": 11.962264150943396, | |
| "grad_norm": 0.04655018821358681, | |
| "learning_rate": 2.170205989174698e-05, | |
| "loss": 0.0132, | |
| "step": 3804 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9480645161290323, | |
| "eval_loss": 0.01816466823220253, | |
| "eval_runtime": 5.4333, | |
| "eval_samples_per_second": 570.554, | |
| "eval_steps_per_second": 11.963, | |
| "step": 3816 | |
| }, | |
| { | |
| "epoch": 12.959119496855346, | |
| "grad_norm": 0.0534706749022007, | |
| "learning_rate": 9.979592497525249e-06, | |
| "loss": 0.0128, | |
| "step": 4121 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.947741935483871, | |
| "eval_loss": 0.01782587543129921, | |
| "eval_runtime": 5.4477, | |
| "eval_samples_per_second": 569.044, | |
| "eval_steps_per_second": 11.932, | |
| "step": 4134 | |
| }, | |
| { | |
| "epoch": 13.955974842767295, | |
| "grad_norm": 0.04927356168627739, | |
| "learning_rate": 2.641137124340782e-06, | |
| "loss": 0.0126, | |
| "step": 4438 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9487096774193549, | |
| "eval_loss": 0.017633505165576935, | |
| "eval_runtime": 5.4206, | |
| "eval_samples_per_second": 571.896, | |
| "eval_steps_per_second": 11.991, | |
| "step": 4452 | |
| }, | |
| { | |
| "epoch": 14.952830188679245, | |
| "grad_norm": 0.049865659326314926, | |
| "learning_rate": 5.412835654343996e-09, | |
| "loss": 0.0125, | |
| "step": 4755 | |
| } | |
| ], | |
| "logging_steps": 317, | |
| "max_steps": 4770, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "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": 1236796014635076.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.18249968951489548, | |
| "learning_rate": 0.0002218415338106894, | |
| "lr_scheduler_type": "cosine", | |
| "num_train_epochs": 15, | |
| "temperature": 8.914270069493725, | |
| "weight_decay": 0.05600918336516608 | |
| } | |
| } | |