| { | |
| "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": 0.6738709677419354, | |
| "eval_loss": 0.40543094277381897, | |
| "eval_runtime": 1.4344, | |
| "eval_samples_per_second": 2161.221, | |
| "eval_steps_per_second": 45.316, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 1.685534591194969e-05, | |
| "loss": 0.6402, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8451612903225807, | |
| "eval_loss": 0.1374775767326355, | |
| "eval_runtime": 1.439, | |
| "eval_samples_per_second": 2154.333, | |
| "eval_steps_per_second": 45.171, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8964516129032258, | |
| "eval_loss": 0.06959949433803558, | |
| "eval_runtime": 1.4502, | |
| "eval_samples_per_second": 2137.661, | |
| "eval_steps_per_second": 44.822, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 1.371069182389937e-05, | |
| "loss": 0.1606, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9161290322580645, | |
| "eval_loss": 0.05019702389836311, | |
| "eval_runtime": 1.46, | |
| "eval_samples_per_second": 2123.327, | |
| "eval_steps_per_second": 44.521, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 1.0566037735849058e-05, | |
| "loss": 0.0795, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9251612903225807, | |
| "eval_loss": 0.0416053868830204, | |
| "eval_runtime": 1.4497, | |
| "eval_samples_per_second": 2138.309, | |
| "eval_steps_per_second": 44.836, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9287096774193548, | |
| "eval_loss": 0.03693937882781029, | |
| "eval_runtime": 1.4473, | |
| "eval_samples_per_second": 2141.876, | |
| "eval_steps_per_second": 44.91, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 7.421383647798742e-06, | |
| "loss": 0.06, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9306451612903226, | |
| "eval_loss": 0.034136172384023666, | |
| "eval_runtime": 1.456, | |
| "eval_samples_per_second": 2129.124, | |
| "eval_steps_per_second": 44.643, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 4.276729559748428e-06, | |
| "loss": 0.0515, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 3180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 650483488638228.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6358446572550642, | |
| "num_train_epochs": 10, | |
| "temperature": 2 | |
| } | |
| } | |