| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9987688519544475, | |
| "eval_steps": 1000, | |
| "global_step": 9741, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0003079765937788728, | |
| "grad_norm": 5.474523544311523, | |
| "learning_rate": 0.0, | |
| "loss": 2.8435, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.3079765937788728, | |
| "grad_norm": 3.9317052364349365, | |
| "learning_rate": 2.991786447638604e-05, | |
| "loss": 2.0411, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3079765937788728, | |
| "eval_cosine_accuracy": 0.9618675112724304, | |
| "eval_loss": 0.6112217903137207, | |
| "eval_runtime": 27.1137, | |
| "eval_samples_per_second": 349.16, | |
| "eval_steps_per_second": 1.365, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6159531875577456, | |
| "grad_norm": 3.60844087600708, | |
| "learning_rate": 2.649555099247091e-05, | |
| "loss": 1.4493, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.6159531875577456, | |
| "eval_cosine_accuracy": 0.9678884744644165, | |
| "eval_loss": 0.5245885848999023, | |
| "eval_runtime": 25.7333, | |
| "eval_samples_per_second": 367.889, | |
| "eval_steps_per_second": 1.438, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9239297813366184, | |
| "grad_norm": 5.359713554382324, | |
| "learning_rate": 2.3073237508555782e-05, | |
| "loss": 1.1061, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.9239297813366184, | |
| "eval_cosine_accuracy": 0.9701066613197327, | |
| "eval_loss": 0.501575231552124, | |
| "eval_runtime": 26.3777, | |
| "eval_samples_per_second": 358.902, | |
| "eval_steps_per_second": 1.403, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.231763619575254, | |
| "grad_norm": 3.523904323577881, | |
| "learning_rate": 1.9654346338124573e-05, | |
| "loss": 1.0082, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.231763619575254, | |
| "eval_cosine_accuracy": 0.9738037586212158, | |
| "eval_loss": 0.48313355445861816, | |
| "eval_runtime": 26.4043, | |
| "eval_samples_per_second": 358.541, | |
| "eval_steps_per_second": 1.401, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.5395506309633733, | |
| "grad_norm": 3.3420140743255615, | |
| "learning_rate": 1.6232032854209448e-05, | |
| "loss": 1.022, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.5395506309633733, | |
| "eval_cosine_accuracy": 0.9766557812690735, | |
| "eval_loss": 0.46782490611076355, | |
| "eval_runtime": 26.5962, | |
| "eval_samples_per_second": 355.953, | |
| "eval_steps_per_second": 1.391, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.847337642351493, | |
| "grad_norm": 3.448085308074951, | |
| "learning_rate": 1.280971937029432e-05, | |
| "loss": 0.9815, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.847337642351493, | |
| "eval_cosine_accuracy": 0.9769726395606995, | |
| "eval_loss": 0.4625483453273773, | |
| "eval_runtime": 26.1888, | |
| "eval_samples_per_second": 361.49, | |
| "eval_steps_per_second": 1.413, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.155124653739612, | |
| "grad_norm": 3.5799896717071533, | |
| "learning_rate": 9.387405886379193e-06, | |
| "loss": 0.9354, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.155124653739612, | |
| "eval_cosine_accuracy": 0.9763388633728027, | |
| "eval_loss": 0.46237874031066895, | |
| "eval_runtime": 26.512, | |
| "eval_samples_per_second": 357.084, | |
| "eval_steps_per_second": 1.396, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.4629116651277316, | |
| "grad_norm": 3.588005542755127, | |
| "learning_rate": 5.9685147159479805e-06, | |
| "loss": 0.8937, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.4629116651277316, | |
| "eval_cosine_accuracy": 0.9765501022338867, | |
| "eval_loss": 0.4602471590042114, | |
| "eval_runtime": 26.1829, | |
| "eval_samples_per_second": 361.572, | |
| "eval_steps_per_second": 1.413, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.7706986765158508, | |
| "grad_norm": 5.595367431640625, | |
| "learning_rate": 2.5462012320328544e-06, | |
| "loss": 0.8904, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.7706986765158508, | |
| "eval_cosine_accuracy": 0.9763388633728027, | |
| "eval_loss": 0.45669785141944885, | |
| "eval_runtime": 26.3088, | |
| "eval_samples_per_second": 359.842, | |
| "eval_steps_per_second": 1.406, | |
| "step": 9000 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 9741, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |