| { | |
| "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": 1.7784031629562378, | |
| "learning_rate": 1.8466376390904693e-05, | |
| "loss": 1.7687, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.67, | |
| "eval_loss": 0.9660505652427673, | |
| "eval_runtime": 1.6131, | |
| "eval_samples_per_second": 1921.733, | |
| "eval_steps_per_second": 40.294, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 1.4204388856887817, | |
| "learning_rate": 1.6932752781809388e-05, | |
| "loss": 0.6811, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.83, | |
| "eval_loss": 0.2671310305595398, | |
| "eval_runtime": 1.5805, | |
| "eval_samples_per_second": 1961.395, | |
| "eval_steps_per_second": 41.126, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 1.0138155221939087, | |
| "learning_rate": 1.539912917271408e-05, | |
| "loss": 0.2272, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9067741935483871, | |
| "eval_loss": 0.11616384238004684, | |
| "eval_runtime": 1.5765, | |
| "eval_samples_per_second": 1966.361, | |
| "eval_steps_per_second": 41.23, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.9868770837783813, | |
| "learning_rate": 1.3865505563618772e-05, | |
| "loss": 0.1117, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9245161290322581, | |
| "eval_loss": 0.07967381924390793, | |
| "eval_runtime": 1.6048, | |
| "eval_samples_per_second": 1931.717, | |
| "eval_steps_per_second": 40.504, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.5568020939826965, | |
| "learning_rate": 1.2331881954523466e-05, | |
| "loss": 0.0762, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9332258064516129, | |
| "eval_loss": 0.0637175664305687, | |
| "eval_runtime": 1.6245, | |
| "eval_samples_per_second": 1908.268, | |
| "eval_steps_per_second": 40.012, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.9021977782249451, | |
| "learning_rate": 1.0798258345428159e-05, | |
| "loss": 0.0609, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9341935483870968, | |
| "eval_loss": 0.05572959780693054, | |
| "eval_runtime": 1.5863, | |
| "eval_samples_per_second": 1954.213, | |
| "eval_steps_per_second": 40.975, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.6561667323112488, | |
| "learning_rate": 9.26463473633285e-06, | |
| "loss": 0.0529, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9380645161290323, | |
| "eval_loss": 0.05086366832256317, | |
| "eval_runtime": 1.5942, | |
| "eval_samples_per_second": 1944.608, | |
| "eval_steps_per_second": 40.774, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.39448273181915283, | |
| "learning_rate": 7.731011127237543e-06, | |
| "loss": 0.0471, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9390322580645162, | |
| "eval_loss": 0.04851401969790459, | |
| "eval_runtime": 1.5838, | |
| "eval_samples_per_second": 1957.35, | |
| "eval_steps_per_second": 41.041, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.562044620513916, | |
| "learning_rate": 6.197387518142236e-06, | |
| "loss": 0.0439, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9345161290322581, | |
| "eval_loss": 0.04487784206867218, | |
| "eval_runtime": 1.5718, | |
| "eval_samples_per_second": 1972.247, | |
| "eval_steps_per_second": 41.354, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.3506462275981903, | |
| "learning_rate": 4.6637639090469286e-06, | |
| "loss": 0.0413, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9396774193548387, | |
| "eval_loss": 0.043323710560798645, | |
| "eval_runtime": 1.5787, | |
| "eval_samples_per_second": 1963.58, | |
| "eval_steps_per_second": 41.172, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 10.965408805031446, | |
| "grad_norm": 0.33104729652404785, | |
| "learning_rate": 3.1301402999516213e-06, | |
| "loss": 0.0394, | |
| "step": 3487 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9396774193548387, | |
| "eval_loss": 0.04240846261382103, | |
| "eval_runtime": 1.5874, | |
| "eval_samples_per_second": 1952.868, | |
| "eval_steps_per_second": 40.947, | |
| "step": 3498 | |
| }, | |
| { | |
| "epoch": 11.962264150943396, | |
| "grad_norm": 0.37558791041374207, | |
| "learning_rate": 1.5965166908563137e-06, | |
| "loss": 0.0382, | |
| "step": 3804 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9403225806451613, | |
| "eval_loss": 0.04175195470452309, | |
| "eval_runtime": 1.593, | |
| "eval_samples_per_second": 1945.992, | |
| "eval_steps_per_second": 40.803, | |
| "step": 3816 | |
| }, | |
| { | |
| "epoch": 12.959119496855346, | |
| "grad_norm": 0.4860643446445465, | |
| "learning_rate": 6.289308176100629e-08, | |
| "loss": 0.0376, | |
| "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": 1072259767520340.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.3858447457522741, | |
| "num_train_epochs": 13, | |
| "temperature": 1 | |
| } | |
| } | |