| { | |
| "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": 0.6109677419354839, | |
| "eval_loss": 0.22998438775539398, | |
| "eval_runtime": 5.4105, | |
| "eval_samples_per_second": 572.959, | |
| "eval_steps_per_second": 12.014, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.5723270440251573, | |
| "grad_norm": 0.5745335221290588, | |
| "learning_rate": 1.650593990216632e-05, | |
| "loss": 0.3689, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8380645161290322, | |
| "eval_loss": 0.10470506548881531, | |
| "eval_runtime": 5.6445, | |
| "eval_samples_per_second": 549.208, | |
| "eval_steps_per_second": 11.516, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.89, | |
| "eval_loss": 0.0663980022072792, | |
| "eval_runtime": 5.6639, | |
| "eval_samples_per_second": 547.33, | |
| "eval_steps_per_second": 11.476, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.1446540880503147, | |
| "grad_norm": 0.49224022030830383, | |
| "learning_rate": 1.3011879804332637e-05, | |
| "loss": 0.1222, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9029032258064517, | |
| "eval_loss": 0.04973406717181206, | |
| "eval_runtime": 5.4039, | |
| "eval_samples_per_second": 573.657, | |
| "eval_steps_per_second": 12.028, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.716981132075472, | |
| "grad_norm": 0.329182505607605, | |
| "learning_rate": 9.517819706498952e-06, | |
| "loss": 0.0746, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9106451612903226, | |
| "eval_loss": 0.0401911735534668, | |
| "eval_runtime": 5.5565, | |
| "eval_samples_per_second": 557.905, | |
| "eval_steps_per_second": 11.698, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9196774193548387, | |
| "eval_loss": 0.03513346612453461, | |
| "eval_runtime": 5.3402, | |
| "eval_samples_per_second": 580.499, | |
| "eval_steps_per_second": 12.172, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.289308176100629, | |
| "grad_norm": 0.28364434838294983, | |
| "learning_rate": 6.02375960866527e-06, | |
| "loss": 0.0576, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2862, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 520991326672152.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.37372962575568247, | |
| "num_train_epochs": 9, | |
| "temperature": 5 | |
| } | |
| } | |