| { | |
| "best_metric": 0.7080158591270447, | |
| "best_model_checkpoint": "final_models/transformer_base_final_2/finetune/mnli/checkpoint-6000", | |
| "epoch": 3.140877598152425, | |
| "global_step": 6800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18, | |
| "eval_accuracy": 0.5156964063644409, | |
| "eval_loss": 0.9809399247169495, | |
| "eval_runtime": 9.0348, | |
| "eval_samples_per_second": 726.302, | |
| "eval_steps_per_second": 90.871, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.884526558891455e-05, | |
| "loss": 1.0588, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.571014940738678, | |
| "eval_loss": 0.9007604718208313, | |
| "eval_runtime": 8.9421, | |
| "eval_samples_per_second": 733.831, | |
| "eval_steps_per_second": 91.813, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.7690531177829104e-05, | |
| "loss": 0.9283, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_accuracy": 0.6338006854057312, | |
| "eval_loss": 0.8218390941619873, | |
| "eval_runtime": 8.965, | |
| "eval_samples_per_second": 731.959, | |
| "eval_steps_per_second": 91.579, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 4.653579676674365e-05, | |
| "loss": 0.836, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.6562023758888245, | |
| "eval_loss": 0.7788540124893188, | |
| "eval_runtime": 8.9831, | |
| "eval_samples_per_second": 730.487, | |
| "eval_steps_per_second": 91.394, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.53810623556582e-05, | |
| "loss": 0.7876, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_accuracy": 0.6472111940383911, | |
| "eval_loss": 0.7989409565925598, | |
| "eval_runtime": 9.0636, | |
| "eval_samples_per_second": 723.995, | |
| "eval_steps_per_second": 90.582, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_accuracy": 0.6717464327812195, | |
| "eval_loss": 0.7579967379570007, | |
| "eval_runtime": 8.9628, | |
| "eval_samples_per_second": 732.136, | |
| "eval_steps_per_second": 91.601, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 4.422632794457275e-05, | |
| "loss": 0.7301, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_accuracy": 0.6728131771087646, | |
| "eval_loss": 0.7662198543548584, | |
| "eval_runtime": 8.9967, | |
| "eval_samples_per_second": 729.376, | |
| "eval_steps_per_second": 91.255, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 4.30715935334873e-05, | |
| "loss": 0.6933, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_accuracy": 0.6782993078231812, | |
| "eval_loss": 0.7824717164039612, | |
| "eval_runtime": 8.9469, | |
| "eval_samples_per_second": 733.442, | |
| "eval_steps_per_second": 91.764, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.1916859122401844e-05, | |
| "loss": 0.6844, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "eval_accuracy": 0.6924718022346497, | |
| "eval_loss": 0.727609395980835, | |
| "eval_runtime": 9.0458, | |
| "eval_samples_per_second": 725.417, | |
| "eval_steps_per_second": 90.76, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 4.07621247113164e-05, | |
| "loss": 0.6825, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_accuracy": 0.6773849725723267, | |
| "eval_loss": 0.7659159302711487, | |
| "eval_runtime": 9.0209, | |
| "eval_samples_per_second": 727.419, | |
| "eval_steps_per_second": 91.011, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "eval_accuracy": 0.6930813789367676, | |
| "eval_loss": 0.7922490239143372, | |
| "eval_runtime": 8.9692, | |
| "eval_samples_per_second": 731.612, | |
| "eval_steps_per_second": 91.535, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 3.960739030023095e-05, | |
| "loss": 0.639, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "eval_accuracy": 0.7077110409736633, | |
| "eval_loss": 0.7321968674659729, | |
| "eval_runtime": 8.9855, | |
| "eval_samples_per_second": 730.29, | |
| "eval_steps_per_second": 91.37, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 3.84526558891455e-05, | |
| "loss": 0.5802, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_accuracy": 0.6982627511024475, | |
| "eval_loss": 0.7771458029747009, | |
| "eval_runtime": 8.9398, | |
| "eval_samples_per_second": 734.024, | |
| "eval_steps_per_second": 91.837, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 3.729792147806005e-05, | |
| "loss": 0.5786, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "eval_accuracy": 0.7005485892295837, | |
| "eval_loss": 0.7566787004470825, | |
| "eval_runtime": 8.9729, | |
| "eval_samples_per_second": 731.313, | |
| "eval_steps_per_second": 91.498, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 3.61431870669746e-05, | |
| "loss": 0.5796, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_accuracy": 0.7080158591270447, | |
| "eval_loss": 0.7298412919044495, | |
| "eval_runtime": 8.9509, | |
| "eval_samples_per_second": 733.11, | |
| "eval_steps_per_second": 91.723, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_accuracy": 0.7008534073829651, | |
| "eval_loss": 0.7494065165519714, | |
| "eval_runtime": 9.0706, | |
| "eval_samples_per_second": 723.435, | |
| "eval_steps_per_second": 90.512, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.498845265588915e-05, | |
| "loss": 0.5797, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "eval_accuracy": 0.7040536403656006, | |
| "eval_loss": 0.8426190614700317, | |
| "eval_runtime": 8.9718, | |
| "eval_samples_per_second": 731.403, | |
| "eval_steps_per_second": 91.509, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "step": 6800, | |
| "total_flos": 5.369124305668608e+16, | |
| "train_loss": 0.7091084985172047, | |
| "train_runtime": 2839.0059, | |
| "train_samples_per_second": 915.039, | |
| "train_steps_per_second": 7.626 | |
| } | |
| ], | |
| "max_steps": 21650, | |
| "num_train_epochs": 10, | |
| "total_flos": 5.369124305668608e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |