| { | |
| "best_metric": 0.4778192639350891, | |
| "best_model_checkpoint": "./vit-lr-0.0001/checkpoint-642", | |
| "epoch": 12.0, | |
| "eval_steps": 500, | |
| "global_step": 3852, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.167571067810059, | |
| "learning_rate": 9.997548148581541e-05, | |
| "loss": 0.6718, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8242024965325936, | |
| "eval_f1": 0.8162630829948523, | |
| "eval_loss": 0.4934074282646179, | |
| "eval_precision": 0.8254417232097437, | |
| "eval_recall": 0.8242024965325936, | |
| "eval_runtime": 37.6292, | |
| "eval_samples_per_second": 76.643, | |
| "eval_steps_per_second": 9.594, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 6.953319072723389, | |
| "learning_rate": 9.990164344447551e-05, | |
| "loss": 0.4045, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8311373092926491, | |
| "eval_f1": 0.8268544179848872, | |
| "eval_loss": 0.4778192639350891, | |
| "eval_precision": 0.8433264041291242, | |
| "eval_recall": 0.8311373092926491, | |
| "eval_runtime": 37.5244, | |
| "eval_samples_per_second": 76.857, | |
| "eval_steps_per_second": 9.62, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 7.354413986206055, | |
| "learning_rate": 9.977855850575433e-05, | |
| "loss": 0.2419, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.840499306518724, | |
| "eval_f1": 0.8409562025706393, | |
| "eval_loss": 0.5132780075073242, | |
| "eval_precision": 0.8524873378932092, | |
| "eval_recall": 0.840499306518724, | |
| "eval_runtime": 36.8758, | |
| "eval_samples_per_second": 78.209, | |
| "eval_steps_per_second": 9.79, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.17630255222320557, | |
| "learning_rate": 9.960634813962617e-05, | |
| "loss": 0.1267, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8495145631067961, | |
| "eval_f1": 0.8447806876150558, | |
| "eval_loss": 0.6153807044029236, | |
| "eval_precision": 0.8490665425851771, | |
| "eval_recall": 0.8495145631067961, | |
| "eval_runtime": 37.1161, | |
| "eval_samples_per_second": 77.702, | |
| "eval_steps_per_second": 9.726, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5389394760131836, | |
| "learning_rate": 9.938518229693118e-05, | |
| "loss": 0.0733, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8422330097087378, | |
| "eval_f1": 0.8361023278787632, | |
| "eval_loss": 0.7844527959823608, | |
| "eval_precision": 0.8420613182520028, | |
| "eval_recall": 0.8422330097087378, | |
| "eval_runtime": 37.3092, | |
| "eval_samples_per_second": 77.3, | |
| "eval_steps_per_second": 9.676, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 24.989282608032227, | |
| "learning_rate": 9.911527924165445e-05, | |
| "loss": 0.0446, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8470873786407767, | |
| "eval_f1": 0.840751584936853, | |
| "eval_loss": 0.8766722679138184, | |
| "eval_precision": 0.8407202080357687, | |
| "eval_recall": 0.8470873786407767, | |
| "eval_runtime": 37.9466, | |
| "eval_samples_per_second": 76.002, | |
| "eval_steps_per_second": 9.513, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.16091497242450714, | |
| "learning_rate": 9.879690533552573e-05, | |
| "loss": 0.0523, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.84500693481276, | |
| "eval_f1": 0.8406651419661996, | |
| "eval_loss": 0.8674203753471375, | |
| "eval_precision": 0.8491745633887275, | |
| "eval_recall": 0.84500693481276, | |
| "eval_runtime": 37.6505, | |
| "eval_samples_per_second": 76.599, | |
| "eval_steps_per_second": 9.588, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.0028496491722762585, | |
| "learning_rate": 9.843159102999166e-05, | |
| "loss": 0.0388, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8398058252427184, | |
| "eval_f1": 0.8387317635385871, | |
| "eval_loss": 0.9753792881965637, | |
| "eval_precision": 0.8565844319804308, | |
| "eval_recall": 0.8398058252427184, | |
| "eval_runtime": 37.8044, | |
| "eval_samples_per_second": 76.287, | |
| "eval_steps_per_second": 9.549, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.0019187598954886198, | |
| "learning_rate": 9.801741382013225e-05, | |
| "loss": 0.0402, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8491678224687933, | |
| "eval_f1": 0.8461066193854649, | |
| "eval_loss": 0.9369620084762573, | |
| "eval_precision": 0.8547133286785668, | |
| "eval_recall": 0.8491678224687933, | |
| "eval_runtime": 37.5745, | |
| "eval_samples_per_second": 76.754, | |
| "eval_steps_per_second": 9.608, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.00408038217574358, | |
| "learning_rate": 9.755584922004499e-05, | |
| "loss": 0.0283, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8509015256588072, | |
| "eval_f1": 0.8482666902922076, | |
| "eval_loss": 0.9217829704284668, | |
| "eval_precision": 0.8495520865912046, | |
| "eval_recall": 0.8509015256588072, | |
| "eval_runtime": 37.6867, | |
| "eval_samples_per_second": 76.526, | |
| "eval_steps_per_second": 9.579, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 2.3583085536956787, | |
| "learning_rate": 9.70473527382648e-05, | |
| "loss": 0.0451, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8474341192787794, | |
| "eval_f1": 0.8401062715209746, | |
| "eval_loss": 0.987198531627655, | |
| "eval_precision": 0.8400167652295873, | |
| "eval_recall": 0.8474341192787794, | |
| "eval_runtime": 37.5611, | |
| "eval_samples_per_second": 76.782, | |
| "eval_steps_per_second": 9.611, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 11.217573165893555, | |
| "learning_rate": 9.649242619942769e-05, | |
| "loss": 0.0549, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8457004160887656, | |
| "eval_f1": 0.8415779150905001, | |
| "eval_loss": 1.0202850103378296, | |
| "eval_precision": 0.8487681253829549, | |
| "eval_recall": 0.8457004160887656, | |
| "eval_runtime": 37.4542, | |
| "eval_samples_per_second": 77.001, | |
| "eval_steps_per_second": 9.638, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "step": 3852, | |
| "total_flos": 4.768760767819088e+18, | |
| "train_loss": 0.15186076372211968, | |
| "train_runtime": 1747.1919, | |
| "train_samples_per_second": 293.5, | |
| "train_steps_per_second": 18.372 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 32100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 4.768760767819088e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |