| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 11.0, | |
| "eval_steps": 500, | |
| "global_step": 3498, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9968553459119497, | |
| "grad_norm": 0.429341197013855, | |
| "learning_rate": 0.00013794880138427262, | |
| "loss": 0.3323, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8664516129032258, | |
| "eval_loss": 0.07744947820901871, | |
| "eval_runtime": 2.7326, | |
| "eval_samples_per_second": 1134.466, | |
| "eval_steps_per_second": 23.787, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.2267828732728958, | |
| "learning_rate": 0.00014653290645786366, | |
| "loss": 0.0632, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9290322580645162, | |
| "eval_loss": 0.034380555152893066, | |
| "eval_runtime": 2.6579, | |
| "eval_samples_per_second": 1166.326, | |
| "eval_steps_per_second": 24.455, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.17225514352321625, | |
| "learning_rate": 0.00013031400584782778, | |
| "loss": 0.0332, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.937741935483871, | |
| "eval_loss": 0.025218434631824493, | |
| "eval_runtime": 2.6227, | |
| "eval_samples_per_second": 1181.985, | |
| "eval_steps_per_second": 24.784, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.10644444078207016, | |
| "learning_rate": 0.0001140951052377919, | |
| "loss": 0.0254, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9448387096774193, | |
| "eval_loss": 0.02074083872139454, | |
| "eval_runtime": 2.6543, | |
| "eval_samples_per_second": 1167.909, | |
| "eval_steps_per_second": 24.488, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.08869536221027374, | |
| "learning_rate": 9.787620462775599e-05, | |
| "loss": 0.0223, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9412903225806452, | |
| "eval_loss": 0.020231781527400017, | |
| "eval_runtime": 2.6557, | |
| "eval_samples_per_second": 1167.32, | |
| "eval_steps_per_second": 24.476, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.08641496300697327, | |
| "learning_rate": 8.16573040177201e-05, | |
| "loss": 0.0201, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9441935483870968, | |
| "eval_loss": 0.01878916658461094, | |
| "eval_runtime": 2.635, | |
| "eval_samples_per_second": 1176.481, | |
| "eval_steps_per_second": 24.668, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.07254982739686966, | |
| "learning_rate": 6.543840340768423e-05, | |
| "loss": 0.0183, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9467741935483871, | |
| "eval_loss": 0.017883770167827606, | |
| "eval_runtime": 2.651, | |
| "eval_samples_per_second": 1169.38, | |
| "eval_steps_per_second": 24.519, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.0713103711605072, | |
| "learning_rate": 4.9219502797648335e-05, | |
| "loss": 0.0171, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9493548387096774, | |
| "eval_loss": 0.017295770347118378, | |
| "eval_runtime": 2.6543, | |
| "eval_samples_per_second": 1167.916, | |
| "eval_steps_per_second": 24.489, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.0737316906452179, | |
| "learning_rate": 3.3000602187612454e-05, | |
| "loss": 0.0162, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9480645161290323, | |
| "eval_loss": 0.01655430532991886, | |
| "eval_runtime": 2.6551, | |
| "eval_samples_per_second": 1167.55, | |
| "eval_steps_per_second": 24.481, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.06824547797441483, | |
| "learning_rate": 1.6781701577576563e-05, | |
| "loss": 0.0155, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.947741935483871, | |
| "eval_loss": 0.016286808997392654, | |
| "eval_runtime": 2.6473, | |
| "eval_samples_per_second": 1170.992, | |
| "eval_steps_per_second": 24.553, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 10.965408805031446, | |
| "grad_norm": 0.058784905821084976, | |
| "learning_rate": 5.628009675406775e-07, | |
| "loss": 0.0149, | |
| "step": 3487 | |
| } | |
| ], | |
| "logging_steps": 317, | |
| "max_steps": 3498, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 11, | |
| "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": 907183422801936.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.5336013283534605, | |
| "fp16": true, | |
| "learning_rate": 0.00016014245712748368, | |
| "lr_scheduler": "cosine", | |
| "num_train_epochs": 11, | |
| "temperature": 9, | |
| "warmup_steps": 368, | |
| "weight_decay": 0.06096677657018104 | |
| } | |
| } | |