| { | |
| "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": 0.5767741935483871, | |
| "eval_loss": 0.18995507061481476, | |
| "eval_runtime": 5.3293, | |
| "eval_samples_per_second": 581.689, | |
| "eval_steps_per_second": 12.197, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.5723270440251573, | |
| "grad_norm": 0.5155988335609436, | |
| "learning_rate": 1.650593990216632e-05, | |
| "loss": 0.3062, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8174193548387096, | |
| "eval_loss": 0.09330115467309952, | |
| "eval_runtime": 5.6232, | |
| "eval_samples_per_second": 551.287, | |
| "eval_steps_per_second": 11.559, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8764516129032258, | |
| "eval_loss": 0.06318038702011108, | |
| "eval_runtime": 5.4039, | |
| "eval_samples_per_second": 573.66, | |
| "eval_steps_per_second": 12.028, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.1446540880503147, | |
| "grad_norm": 0.42988139390945435, | |
| "learning_rate": 1.3011879804332637e-05, | |
| "loss": 0.108, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8990322580645161, | |
| "eval_loss": 0.04927925392985344, | |
| "eval_runtime": 5.4126, | |
| "eval_samples_per_second": 572.741, | |
| "eval_steps_per_second": 12.009, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.716981132075472, | |
| "grad_norm": 0.297043114900589, | |
| "learning_rate": 9.517819706498952e-06, | |
| "loss": 0.0699, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9074193548387097, | |
| "eval_loss": 0.040702205151319504, | |
| "eval_runtime": 5.5616, | |
| "eval_samples_per_second": 557.389, | |
| "eval_steps_per_second": 11.687, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9103225806451613, | |
| "eval_loss": 0.035840049386024475, | |
| "eval_runtime": 5.3906, | |
| "eval_samples_per_second": 575.073, | |
| "eval_steps_per_second": 12.058, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.289308176100629, | |
| "grad_norm": 0.26027998328208923, | |
| "learning_rate": 6.02375960866527e-06, | |
| "loss": 0.0552, | |
| "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": 0.39616967623231003, | |
| "num_train_epochs": 9, | |
| "temperature": 20 | |
| } | |
| } | |