| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 4134, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9968553459119497, | |
| "grad_norm": 0.3114970326423645, | |
| "learning_rate": 0.00018516884358804785, | |
| "loss": 0.285, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8787096774193548, | |
| "eval_loss": 0.06010065972805023, | |
| "eval_runtime": 2.5938, | |
| "eval_samples_per_second": 1195.147, | |
| "eval_steps_per_second": 25.06, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.2971077263355255, | |
| "learning_rate": 0.00016979066087455266, | |
| "loss": 0.0516, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9296774193548387, | |
| "eval_loss": 0.03248392790555954, | |
| "eval_runtime": 2.6049, | |
| "eval_samples_per_second": 1190.073, | |
| "eval_steps_per_second": 24.953, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.24367552995681763, | |
| "learning_rate": 0.00015441247816105745, | |
| "loss": 0.0306, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9354838709677419, | |
| "eval_loss": 0.02789517492055893, | |
| "eval_runtime": 2.6186, | |
| "eval_samples_per_second": 1183.836, | |
| "eval_steps_per_second": 24.822, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.1115935891866684, | |
| "learning_rate": 0.00013903429544756226, | |
| "loss": 0.0247, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9432258064516129, | |
| "eval_loss": 0.022768640890717506, | |
| "eval_runtime": 2.6313, | |
| "eval_samples_per_second": 1178.135, | |
| "eval_steps_per_second": 24.703, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.09492151439189911, | |
| "learning_rate": 0.00012365611273406705, | |
| "loss": 0.0211, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9416129032258065, | |
| "eval_loss": 0.021043915301561356, | |
| "eval_runtime": 2.6325, | |
| "eval_samples_per_second": 1177.607, | |
| "eval_steps_per_second": 24.692, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.1175580695271492, | |
| "learning_rate": 0.00010827793002057186, | |
| "loss": 0.0195, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9429032258064516, | |
| "eval_loss": 0.020259153097867966, | |
| "eval_runtime": 2.6508, | |
| "eval_samples_per_second": 1169.473, | |
| "eval_steps_per_second": 24.521, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.11239363253116608, | |
| "learning_rate": 9.289974730707666e-05, | |
| "loss": 0.0179, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9454838709677419, | |
| "eval_loss": 0.01921679638326168, | |
| "eval_runtime": 2.6313, | |
| "eval_samples_per_second": 1178.137, | |
| "eval_steps_per_second": 24.703, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.06852811574935913, | |
| "learning_rate": 7.752156459358147e-05, | |
| "loss": 0.0166, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.944516129032258, | |
| "eval_loss": 0.01865958236157894, | |
| "eval_runtime": 2.6643, | |
| "eval_samples_per_second": 1163.539, | |
| "eval_steps_per_second": 24.397, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.06738817691802979, | |
| "learning_rate": 6.214338188008627e-05, | |
| "loss": 0.0158, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.944516129032258, | |
| "eval_loss": 0.017864950001239777, | |
| "eval_runtime": 2.6486, | |
| "eval_samples_per_second": 1170.45, | |
| "eval_steps_per_second": 24.542, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.062413159757852554, | |
| "learning_rate": 4.6765199166591074e-05, | |
| "loss": 0.015, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9467741935483871, | |
| "eval_loss": 0.017261695116758347, | |
| "eval_runtime": 2.652, | |
| "eval_samples_per_second": 1168.93, | |
| "eval_steps_per_second": 24.51, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 10.965408805031446, | |
| "grad_norm": 0.05933304503560066, | |
| "learning_rate": 3.138701645309588e-05, | |
| "loss": 0.0143, | |
| "step": 3487 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9464516129032258, | |
| "eval_loss": 0.016805030405521393, | |
| "eval_runtime": 2.653, | |
| "eval_samples_per_second": 1168.48, | |
| "eval_steps_per_second": 24.5, | |
| "step": 3498 | |
| }, | |
| { | |
| "epoch": 11.962264150943396, | |
| "grad_norm": 0.05458727478981018, | |
| "learning_rate": 1.6008833739600677e-05, | |
| "loss": 0.0138, | |
| "step": 3804 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9458064516129032, | |
| "eval_loss": 0.016528310254216194, | |
| "eval_runtime": 2.6453, | |
| "eval_samples_per_second": 1171.873, | |
| "eval_steps_per_second": 24.572, | |
| "step": 3816 | |
| }, | |
| { | |
| "epoch": 12.959119496855346, | |
| "grad_norm": 0.05435788258910179, | |
| "learning_rate": 6.306510261054813e-07, | |
| "loss": 0.0133, | |
| "step": 4121 | |
| } | |
| ], | |
| "logging_steps": 317, | |
| "max_steps": 4134, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "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": 1072409708161512.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.3513465070451599, | |
| "fp16": false, | |
| "learning_rate": 0.00018909828459685892, | |
| "lr_scheduler": "cosine", | |
| "num_train_epochs": 13, | |
| "temperature": 9, | |
| "warmup_steps": 236, | |
| "weight_decay": 0.09716219849882443 | |
| } | |
| } | |