| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.861635220125786, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6061290322580645, | |
| "eval_loss": 0.2067662924528122, | |
| "eval_runtime": 1.83, | |
| "eval_samples_per_second": 1693.985, | |
| "eval_steps_per_second": 35.519, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 1.650593990216632e-05, | |
| "loss": 0.3335, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8312903225806452, | |
| "eval_loss": 0.09916964918375015, | |
| "eval_runtime": 2.1477, | |
| "eval_samples_per_second": 1443.399, | |
| "eval_steps_per_second": 30.265, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8816129032258064, | |
| "eval_loss": 0.06577017158269882, | |
| "eval_runtime": 1.8521, | |
| "eval_samples_per_second": 1673.812, | |
| "eval_steps_per_second": 35.096, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 1.3011879804332637e-05, | |
| "loss": 0.1145, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8987096774193548, | |
| "eval_loss": 0.04967508092522621, | |
| "eval_runtime": 1.8807, | |
| "eval_samples_per_second": 1648.287, | |
| "eval_steps_per_second": 34.561, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 9.517819706498952e-06, | |
| "loss": 0.0721, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9074193548387097, | |
| "eval_loss": 0.04028124362230301, | |
| "eval_runtime": 1.8504, | |
| "eval_samples_per_second": 1675.341, | |
| "eval_steps_per_second": 35.128, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9187096774193548, | |
| "eval_loss": 0.03542502969503403, | |
| "eval_runtime": 1.8593, | |
| "eval_samples_per_second": 1667.283, | |
| "eval_steps_per_second": 34.959, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 6.02375960866527e-06, | |
| "loss": 0.0569, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9206451612903226, | |
| "eval_loss": 0.03249173238873482, | |
| "eval_runtime": 2.3133, | |
| "eval_samples_per_second": 1340.088, | |
| "eval_steps_per_second": 28.099, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 2.5296995108315863e-06, | |
| "loss": 0.0496, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2862, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 650483488638228.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.08109385800901192, | |
| "num_train_epochs": 9, | |
| "temperature": 8 | |
| } | |
| } | |