| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 14.0, | |
| "eval_steps": 500, | |
| "global_step": 4452, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9968553459119497, | |
| "grad_norm": 0.3429218530654907, | |
| "learning_rate": 0.00019453796448455724, | |
| "loss": 0.2573, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8848387096774194, | |
| "eval_loss": 0.05875122919678688, | |
| "eval_runtime": 2.6188, | |
| "eval_samples_per_second": 1183.755, | |
| "eval_steps_per_second": 24.821, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.2231501191854477, | |
| "learning_rate": 0.00017962417131851016, | |
| "loss": 0.0475, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9248387096774193, | |
| "eval_loss": 0.033762380480766296, | |
| "eval_runtime": 2.6039, | |
| "eval_samples_per_second": 1190.545, | |
| "eval_steps_per_second": 24.963, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.20701543986797333, | |
| "learning_rate": 0.00016471037815246308, | |
| "loss": 0.0305, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9387096774193548, | |
| "eval_loss": 0.025485960766673088, | |
| "eval_runtime": 2.6072, | |
| "eval_samples_per_second": 1189.026, | |
| "eval_steps_per_second": 24.931, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.2078278660774231, | |
| "learning_rate": 0.00014979658498641602, | |
| "loss": 0.0242, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9412903225806452, | |
| "eval_loss": 0.024921948090195656, | |
| "eval_runtime": 2.6163, | |
| "eval_samples_per_second": 1184.865, | |
| "eval_steps_per_second": 24.844, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.09209390729665756, | |
| "learning_rate": 0.00013488279182036894, | |
| "loss": 0.0213, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9454838709677419, | |
| "eval_loss": 0.021658778190612793, | |
| "eval_runtime": 2.6077, | |
| "eval_samples_per_second": 1188.789, | |
| "eval_steps_per_second": 24.926, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.07631496340036392, | |
| "learning_rate": 0.00011996899865432188, | |
| "loss": 0.0191, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9435483870967742, | |
| "eval_loss": 0.021196924149990082, | |
| "eval_runtime": 2.6236, | |
| "eval_samples_per_second": 1181.597, | |
| "eval_steps_per_second": 24.775, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.06425660103559494, | |
| "learning_rate": 0.0001050552054882748, | |
| "loss": 0.0176, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9493548387096774, | |
| "eval_loss": 0.019918320700526237, | |
| "eval_runtime": 2.6075, | |
| "eval_samples_per_second": 1188.867, | |
| "eval_steps_per_second": 24.928, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.0663381814956665, | |
| "learning_rate": 9.014141232222773e-05, | |
| "loss": 0.0165, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9490322580645161, | |
| "eval_loss": 0.019041651859879494, | |
| "eval_runtime": 2.6053, | |
| "eval_samples_per_second": 1189.869, | |
| "eval_steps_per_second": 24.949, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.0676058828830719, | |
| "learning_rate": 7.522761915618066e-05, | |
| "loss": 0.0156, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9487096774193549, | |
| "eval_loss": 0.018456541001796722, | |
| "eval_runtime": 2.6235, | |
| "eval_samples_per_second": 1181.605, | |
| "eval_steps_per_second": 24.776, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.062172286212444305, | |
| "learning_rate": 6.031382599013358e-05, | |
| "loss": 0.0148, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.947741935483871, | |
| "eval_loss": 0.01798408292233944, | |
| "eval_runtime": 2.6268, | |
| "eval_samples_per_second": 1180.13, | |
| "eval_steps_per_second": 24.745, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 10.965408805031446, | |
| "grad_norm": 0.0584622323513031, | |
| "learning_rate": 4.5400032824086506e-05, | |
| "loss": 0.0141, | |
| "step": 3487 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9503225806451613, | |
| "eval_loss": 0.017367394641041756, | |
| "eval_runtime": 2.6152, | |
| "eval_samples_per_second": 1185.368, | |
| "eval_steps_per_second": 24.854, | |
| "step": 3498 | |
| }, | |
| { | |
| "epoch": 11.962264150943396, | |
| "grad_norm": 0.05319051072001457, | |
| "learning_rate": 3.048623965803944e-05, | |
| "loss": 0.0135, | |
| "step": 3804 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9493548387096774, | |
| "eval_loss": 0.017015676945447922, | |
| "eval_runtime": 2.5922, | |
| "eval_samples_per_second": 1195.915, | |
| "eval_steps_per_second": 25.076, | |
| "step": 3816 | |
| }, | |
| { | |
| "epoch": 12.959119496855346, | |
| "grad_norm": 0.05101251229643822, | |
| "learning_rate": 1.5572446491992368e-05, | |
| "loss": 0.013, | |
| "step": 4121 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9493548387096774, | |
| "eval_loss": 0.016792386770248413, | |
| "eval_runtime": 2.6169, | |
| "eval_samples_per_second": 1184.591, | |
| "eval_steps_per_second": 24.838, | |
| "step": 4134 | |
| }, | |
| { | |
| "epoch": 13.955974842767295, | |
| "grad_norm": 0.04865499958395958, | |
| "learning_rate": 6.586533259452965e-07, | |
| "loss": 0.0126, | |
| "step": 4438 | |
| } | |
| ], | |
| "logging_steps": 317, | |
| "max_steps": 4452, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 14, | |
| "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": 1154143441198440.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.34251596991370203, | |
| "fp16": false, | |
| "learning_rate": 0.00020126563774242704, | |
| "lr_scheduler": "cosine", | |
| "num_train_epochs": 14, | |
| "temperature": 10, | |
| "warmup_steps": 174, | |
| "weight_decay": 0.1646286765476401 | |
| } | |
| } | |