| { | |
| "best_metric": 0.8414872798434442, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-27/checkpoint-768", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 864, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.020627737045288, | |
| "learning_rate": 0.00020460634625664984, | |
| "loss": 0.494, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7886497064579256, | |
| "eval_f1": 0.8032786885245903, | |
| "eval_loss": 0.4278793931007385, | |
| "eval_precision": 0.7512776831345827, | |
| "eval_recall": 0.863013698630137, | |
| "eval_runtime": 24.9337, | |
| "eval_samples_per_second": 40.989, | |
| "eval_steps_per_second": 1.283, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.554121732711792, | |
| "learning_rate": 0.00017903055297456864, | |
| "loss": 0.4258, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8062622309197651, | |
| "eval_f1": 0.8241563055062168, | |
| "eval_loss": 0.40374234318733215, | |
| "eval_precision": 0.7544715447154472, | |
| "eval_recall": 0.9080234833659491, | |
| "eval_runtime": 25.4225, | |
| "eval_samples_per_second": 40.201, | |
| "eval_steps_per_second": 1.259, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.225475311279297, | |
| "learning_rate": 0.00015345475969248738, | |
| "loss": 0.405, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8277886497064579, | |
| "eval_f1": 0.84, | |
| "eval_loss": 0.39554914832115173, | |
| "eval_precision": 0.7843803056027164, | |
| "eval_recall": 0.9041095890410958, | |
| "eval_runtime": 25.5434, | |
| "eval_samples_per_second": 40.01, | |
| "eval_steps_per_second": 1.253, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.6178488731384277, | |
| "learning_rate": 0.00012787896641040618, | |
| "loss": 0.3903, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.821917808219178, | |
| "eval_f1": 0.8222656250000001, | |
| "eval_loss": 0.38986799120903015, | |
| "eval_precision": 0.8206627680311891, | |
| "eval_recall": 0.8238747553816047, | |
| "eval_runtime": 25.4279, | |
| "eval_samples_per_second": 40.192, | |
| "eval_steps_per_second": 1.258, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.1577227115631104, | |
| "learning_rate": 0.00010230317312832492, | |
| "loss": 0.3846, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8228962818003914, | |
| "eval_f1": 0.8309990662931839, | |
| "eval_loss": 0.3861733376979828, | |
| "eval_precision": 0.7946428571428571, | |
| "eval_recall": 0.8708414872798435, | |
| "eval_runtime": 25.7302, | |
| "eval_samples_per_second": 39.72, | |
| "eval_steps_per_second": 1.244, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.0598785877227783, | |
| "learning_rate": 7.672737984624369e-05, | |
| "loss": 0.3784, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8317025440313112, | |
| "eval_f1": 0.8472468916518652, | |
| "eval_loss": 0.3895849585533142, | |
| "eval_precision": 0.775609756097561, | |
| "eval_recall": 0.9334637964774951, | |
| "eval_runtime": 25.139, | |
| "eval_samples_per_second": 40.654, | |
| "eval_steps_per_second": 1.273, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.7429569959640503, | |
| "learning_rate": 5.115158656416246e-05, | |
| "loss": 0.3756, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8405088062622309, | |
| "eval_f1": 0.8530207394048693, | |
| "eval_loss": 0.3818517029285431, | |
| "eval_precision": 0.7909698996655519, | |
| "eval_recall": 0.9256360078277887, | |
| "eval_runtime": 25.2278, | |
| "eval_samples_per_second": 40.511, | |
| "eval_steps_per_second": 1.268, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.883507013320923, | |
| "learning_rate": 2.557579328208123e-05, | |
| "loss": 0.3725, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8414872798434442, | |
| "eval_f1": 0.8527272727272727, | |
| "eval_loss": 0.38016170263290405, | |
| "eval_precision": 0.7962648556876061, | |
| "eval_recall": 0.9178082191780822, | |
| "eval_runtime": 24.8536, | |
| "eval_samples_per_second": 41.121, | |
| "eval_steps_per_second": 1.288, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 2.0533533096313477, | |
| "learning_rate": 0.0, | |
| "loss": 0.3665, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.837573385518591, | |
| "eval_f1": 0.8471454880294659, | |
| "eval_loss": 0.38007956743240356, | |
| "eval_precision": 0.8, | |
| "eval_recall": 0.9001956947162426, | |
| "eval_runtime": 25.3933, | |
| "eval_samples_per_second": 40.247, | |
| "eval_steps_per_second": 1.26, | |
| "step": 864 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 864, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 2121256775520.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.718256826807231, | |
| "learning_rate": 0.0002301821395387311, | |
| "num_train_epochs": 9, | |
| "temperature": 6 | |
| } | |
| } | |