| { | |
| "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": 0.8087115287780762, | |
| "learning_rate": 1.916642112888053e-05, | |
| "loss": 0.6426, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4332258064516129, | |
| "eval_loss": 0.3768153786659241, | |
| "eval_runtime": 2.7098, | |
| "eval_samples_per_second": 1144.007, | |
| "eval_steps_per_second": 23.987, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.5955724120140076, | |
| "learning_rate": 2.817448208700141e-05, | |
| "loss": 0.2402, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8554838709677419, | |
| "eval_loss": 0.09723836928606033, | |
| "eval_runtime": 2.6882, | |
| "eval_samples_per_second": 1153.187, | |
| "eval_steps_per_second": 24.18, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.3765904903411865, | |
| "learning_rate": 2.6015074248128236e-05, | |
| "loss": 0.0949, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9080645161290323, | |
| "eval_loss": 0.04660297930240631, | |
| "eval_runtime": 2.6607, | |
| "eval_samples_per_second": 1165.117, | |
| "eval_steps_per_second": 24.43, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.26715776324272156, | |
| "learning_rate": 2.385566640925506e-05, | |
| "loss": 0.0599, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9283870967741935, | |
| "eval_loss": 0.03252074867486954, | |
| "eval_runtime": 2.6662, | |
| "eval_samples_per_second": 1162.716, | |
| "eval_steps_per_second": 24.38, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.20171727240085602, | |
| "learning_rate": 2.1696258570381886e-05, | |
| "loss": 0.0462, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9309677419354838, | |
| "eval_loss": 0.026936793699860573, | |
| "eval_runtime": 2.6975, | |
| "eval_samples_per_second": 1149.215, | |
| "eval_steps_per_second": 24.096, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.21952152252197266, | |
| "learning_rate": 1.9536850731508715e-05, | |
| "loss": 0.0395, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9345161290322581, | |
| "eval_loss": 0.023858336731791496, | |
| "eval_runtime": 2.6565, | |
| "eval_samples_per_second": 1166.971, | |
| "eval_steps_per_second": 24.469, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.30627548694610596, | |
| "learning_rate": 1.737744289263554e-05, | |
| "loss": 0.0356, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9354838709677419, | |
| "eval_loss": 0.022071754559874535, | |
| "eval_runtime": 2.6587, | |
| "eval_samples_per_second": 1165.985, | |
| "eval_steps_per_second": 24.448, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.16566617786884308, | |
| "learning_rate": 1.5218035053762365e-05, | |
| "loss": 0.0328, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9370967741935484, | |
| "eval_loss": 0.021090181544423103, | |
| "eval_runtime": 2.6745, | |
| "eval_samples_per_second": 1159.074, | |
| "eval_steps_per_second": 24.303, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.16504672169685364, | |
| "learning_rate": 1.3058627214889192e-05, | |
| "loss": 0.0308, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9348387096774193, | |
| "eval_loss": 0.019647156819701195, | |
| "eval_runtime": 2.6634, | |
| "eval_samples_per_second": 1163.916, | |
| "eval_steps_per_second": 24.405, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.1753920316696167, | |
| "learning_rate": 1.0899219376016019e-05, | |
| "loss": 0.0293, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9367741935483871, | |
| "eval_loss": 0.019323358312249184, | |
| "eval_runtime": 2.6766, | |
| "eval_samples_per_second": 1158.172, | |
| "eval_steps_per_second": 24.284, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 10.965408805031446, | |
| "grad_norm": 0.16472382843494415, | |
| "learning_rate": 8.739811537142844e-06, | |
| "loss": 0.028, | |
| "step": 3487 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9338709677419355, | |
| "eval_loss": 0.019101083278656006, | |
| "eval_runtime": 2.6594, | |
| "eval_samples_per_second": 1165.659, | |
| "eval_steps_per_second": 24.441, | |
| "step": 3498 | |
| }, | |
| { | |
| "epoch": 11.962264150943396, | |
| "grad_norm": 0.1276603639125824, | |
| "learning_rate": 6.580403698269671e-06, | |
| "loss": 0.0271, | |
| "step": 3804 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9341935483870968, | |
| "eval_loss": 0.018606653437018394, | |
| "eval_runtime": 2.6418, | |
| "eval_samples_per_second": 1173.436, | |
| "eval_steps_per_second": 24.604, | |
| "step": 3816 | |
| }, | |
| { | |
| "epoch": 12.959119496855346, | |
| "grad_norm": 0.13854487240314484, | |
| "learning_rate": 4.420995859396498e-06, | |
| "loss": 0.0264, | |
| "step": 4121 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9358064516129032, | |
| "eval_loss": 0.018243877217173576, | |
| "eval_runtime": 2.6544, | |
| "eval_samples_per_second": 1167.882, | |
| "eval_steps_per_second": 24.488, | |
| "step": 4134 | |
| }, | |
| { | |
| "epoch": 13.955974842767295, | |
| "grad_norm": 0.1205814927816391, | |
| "learning_rate": 2.2615880205233243e-06, | |
| "loss": 0.0259, | |
| "step": 4438 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9361290322580645, | |
| "eval_loss": 0.01793498359620571, | |
| "eval_runtime": 2.6531, | |
| "eval_samples_per_second": 1168.457, | |
| "eval_steps_per_second": 24.5, | |
| "step": 4452 | |
| }, | |
| { | |
| "epoch": 14.952830188679245, | |
| "grad_norm": 0.13082493841648102, | |
| "learning_rate": 1.0218018165015018e-07, | |
| "loss": 0.0256, | |
| "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": 1259981299661700.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.41037073052052975, | |
| "fp16": false, | |
| "learning_rate": 2.920309591561292e-05, | |
| "lr_scheduler": "cosine", | |
| "num_train_epochs": 15, | |
| "temperature": 4, | |
| "warmup_steps": 483, | |
| "weight_decay": 0.1243517366819557 | |
| } | |
| } | |