| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.289308176100629, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6506451612903226, | |
| "eval_loss": 0.2873533070087433, | |
| "eval_runtime": 5.9654, | |
| "eval_samples_per_second": 519.661, | |
| "eval_steps_per_second": 10.896, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 1.650593990216632e-05, | |
| "loss": 0.4591, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8432258064516129, | |
| "eval_loss": 0.1160515546798706, | |
| "eval_runtime": 6.0854, | |
| "eval_samples_per_second": 509.418, | |
| "eval_steps_per_second": 10.681, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8958064516129032, | |
| "eval_loss": 0.06775005906820297, | |
| "eval_runtime": 6.0485, | |
| "eval_samples_per_second": 512.528, | |
| "eval_steps_per_second": 10.747, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 1.3011879804332637e-05, | |
| "loss": 0.1363, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9106451612903226, | |
| "eval_loss": 0.04885930195450783, | |
| "eval_runtime": 5.9946, | |
| "eval_samples_per_second": 517.134, | |
| "eval_steps_per_second": 10.843, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 9.517819706498952e-06, | |
| "loss": 0.0769, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9196774193548387, | |
| "eval_loss": 0.039541397243738174, | |
| "eval_runtime": 6.0954, | |
| "eval_samples_per_second": 508.58, | |
| "eval_steps_per_second": 10.664, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9267741935483871, | |
| "eval_loss": 0.034905772656202316, | |
| "eval_runtime": 5.9756, | |
| "eval_samples_per_second": 518.776, | |
| "eval_steps_per_second": 10.878, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 6.02375960866527e-06, | |
| "loss": 0.059, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2862, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 697479243717816.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.706898380229718, | |
| "num_train_epochs": 9, | |
| "temperature": 3 | |
| } | |
| } | |