--- license: apache-2.0 base_model: facebook/hubert-base-ls960 tags: - generated_from_trainer metrics: - accuracy - precision - recall - f1 model-index: - name: hubert-classifier-aug-fold-0 results: [] --- # hubert-classifier-aug-fold-0 This model is a fine-tuned version of [facebook/hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.4752 - Accuracy: 0.8895 - Precision: 0.9022 - Recall: 0.8895 - F1: 0.8885 - Binary: 0.9222 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0001 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 128 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 30 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | Binary | |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:------:| | No log | 0.13 | 50 | 3.9894 | 0.0472 | 0.0049 | 0.0472 | 0.0080 | 0.3132 | | No log | 0.27 | 100 | 3.4226 | 0.0970 | 0.0691 | 0.0970 | 0.0477 | 0.3615 | | No log | 0.4 | 150 | 3.1173 | 0.1509 | 0.0772 | 0.1509 | 0.0759 | 0.3985 | | No log | 0.54 | 200 | 2.7885 | 0.2439 | 0.1966 | 0.2439 | 0.1769 | 0.4664 | | No log | 0.67 | 250 | 2.4226 | 0.3477 | 0.2627 | 0.3477 | 0.2654 | 0.5412 | | No log | 0.81 | 300 | 2.0733 | 0.4515 | 0.4074 | 0.4515 | 0.3860 | 0.6143 | | No log | 0.94 | 350 | 1.8531 | 0.5512 | 0.5247 | 0.5512 | 0.4904 | 0.6844 | | 3.0592 | 1.08 | 400 | 1.6649 | 0.5809 | 0.5680 | 0.5809 | 0.5365 | 0.7046 | | 3.0592 | 1.21 | 450 | 1.4351 | 0.6199 | 0.6214 | 0.6199 | 0.5796 | 0.7327 | | 3.0592 | 1.35 | 500 | 1.2309 | 0.6954 | 0.6842 | 0.6954 | 0.6636 | 0.7864 | | 3.0592 | 1.48 | 550 | 1.1706 | 0.7089 | 0.7292 | 0.7089 | 0.6899 | 0.7947 | | 3.0592 | 1.62 | 600 | 1.0363 | 0.7264 | 0.7374 | 0.7264 | 0.7076 | 0.8078 | | 3.0592 | 1.75 | 650 | 0.9977 | 0.7642 | 0.7769 | 0.7642 | 0.7508 | 0.8344 | | 3.0592 | 1.89 | 700 | 0.9239 | 0.7655 | 0.7900 | 0.7655 | 0.7497 | 0.8367 | | 1.4103 | 2.02 | 750 | 0.8664 | 0.7722 | 0.7836 | 0.7722 | 0.7587 | 0.8398 | | 1.4103 | 2.16 | 800 | 0.8351 | 0.7844 | 0.7920 | 0.7844 | 0.7716 | 0.8480 | | 1.4103 | 2.29 | 850 | 0.7602 | 0.7965 | 0.8065 | 0.7965 | 0.7904 | 0.8578 | | 1.4103 | 2.43 | 900 | 0.7829 | 0.8100 | 0.8364 | 0.8100 | 0.8052 | 0.8678 | | 1.4103 | 2.56 | 950 | 0.7197 | 0.8005 | 0.8158 | 0.8005 | 0.7903 | 0.8616 | | 1.4103 | 2.7 | 1000 | 0.7602 | 0.8113 | 0.8297 | 0.8113 | 0.8059 | 0.8683 | | 1.4103 | 2.83 | 1050 | 0.6590 | 0.8342 | 0.8489 | 0.8342 | 0.8266 | 0.8850 | | 1.4103 | 2.97 | 1100 | 0.6178 | 0.8396 | 0.8546 | 0.8396 | 0.8320 | 0.8880 | | 0.8796 | 3.1 | 1150 | 0.6263 | 0.8396 | 0.8577 | 0.8396 | 0.8344 | 0.8871 | | 0.8796 | 3.24 | 1200 | 0.6099 | 0.8369 | 0.8532 | 0.8369 | 0.8302 | 0.8861 | | 0.8796 | 3.37 | 1250 | 0.6266 | 0.8329 | 0.8505 | 0.8329 | 0.8289 | 0.8815 | | 0.8796 | 3.51 | 1300 | 0.5873 | 0.8437 | 0.8564 | 0.8437 | 0.8409 | 0.8896 | | 0.8796 | 3.64 | 1350 | 0.6750 | 0.8396 | 0.8572 | 0.8396 | 0.8359 | 0.8875 | | 0.8796 | 3.78 | 1400 | 0.6247 | 0.8383 | 0.8534 | 0.8383 | 0.8356 | 0.8856 | | 0.8796 | 3.91 | 1450 | 0.6302 | 0.8369 | 0.8483 | 0.8369 | 0.8328 | 0.8852 | | 0.6486 | 4.05 | 1500 | 0.6063 | 0.8558 | 0.8670 | 0.8558 | 0.8504 | 0.9001 | | 0.6486 | 4.18 | 1550 | 0.6152 | 0.8410 | 0.8564 | 0.8410 | 0.8359 | 0.8884 | | 0.6486 | 4.32 | 1600 | 0.6123 | 0.8504 | 0.8653 | 0.8504 | 0.8438 | 0.8949 | | 0.6486 | 4.45 | 1650 | 0.6590 | 0.8464 | 0.8608 | 0.8464 | 0.8410 | 0.8934 | | 0.6486 | 4.59 | 1700 | 0.6929 | 0.8464 | 0.8599 | 0.8464 | 0.8419 | 0.8937 | | 0.6486 | 4.72 | 1750 | 0.5531 | 0.8598 | 0.8753 | 0.8598 | 0.8574 | 0.9020 | | 0.6486 | 4.86 | 1800 | 0.6241 | 0.8504 | 0.8635 | 0.8504 | 0.8446 | 0.8949 | | 0.6486 | 4.99 | 1850 | 0.6102 | 0.8531 | 0.8632 | 0.8531 | 0.8484 | 0.8966 | | 0.5302 | 5.12 | 1900 | 0.6633 | 0.8612 | 0.8764 | 0.8612 | 0.8541 | 0.9026 | | 0.5302 | 5.26 | 1950 | 0.6566 | 0.8531 | 0.8633 | 0.8531 | 0.8489 | 0.8968 | | 0.5302 | 5.39 | 2000 | 0.7052 | 0.8342 | 0.8598 | 0.8342 | 0.8312 | 0.8836 | | 0.5302 | 5.53 | 2050 | 0.6527 | 0.8518 | 0.8618 | 0.8518 | 0.8491 | 0.8958 | | 0.5302 | 5.66 | 2100 | 0.6170 | 0.8598 | 0.8714 | 0.8598 | 0.8558 | 0.9011 | | 0.5302 | 5.8 | 2150 | 0.6832 | 0.8491 | 0.8648 | 0.8491 | 0.8462 | 0.8939 | | 0.5302 | 5.93 | 2200 | 0.6431 | 0.8571 | 0.8733 | 0.8571 | 0.8540 | 0.8992 | | 0.4492 | 6.07 | 2250 | 0.6744 | 0.8437 | 0.8664 | 0.8437 | 0.8408 | 0.8898 | | 0.4492 | 6.2 | 2300 | 0.6106 | 0.8544 | 0.8641 | 0.8544 | 0.8499 | 0.8977 | | 0.4492 | 6.34 | 2350 | 0.5262 | 0.8652 | 0.8794 | 0.8652 | 0.8636 | 0.9053 | | 0.4492 | 6.47 | 2400 | 0.6457 | 0.8544 | 0.8689 | 0.8544 | 0.8519 | 0.8981 | | 0.4492 | 6.61 | 2450 | 0.6413 | 0.8625 | 0.8688 | 0.8625 | 0.8584 | 0.9034 | | 0.4492 | 6.74 | 2500 | 0.6634 | 0.8504 | 0.8634 | 0.8504 | 0.8476 | 0.8941 | | 0.4492 | 6.88 | 2550 | 0.6251 | 0.8612 | 0.8708 | 0.8612 | 0.8560 | 0.9024 | | 0.4054 | 7.01 | 2600 | 0.6368 | 0.8558 | 0.8682 | 0.8558 | 0.8501 | 0.8988 | | 0.4054 | 7.15 | 2650 | 0.6141 | 0.8720 | 0.8825 | 0.8720 | 0.8680 | 0.9097 | | 0.4054 | 7.28 | 2700 | 0.6492 | 0.8558 | 0.8711 | 0.8558 | 0.8531 | 0.8988 | | 0.4054 | 7.42 | 2750 | 0.6126 | 0.8760 | 0.8842 | 0.8760 | 0.8720 | 0.9135 | | 0.4054 | 7.55 | 2800 | 0.6686 | 0.8720 | 0.8827 | 0.8720 | 0.8683 | 0.9089 | | 0.4054 | 7.69 | 2850 | 0.6718 | 0.8652 | 0.8774 | 0.8652 | 0.8629 | 0.9053 | | 0.4054 | 7.82 | 2900 | 0.6315 | 0.8706 | 0.8832 | 0.8706 | 0.8681 | 0.9092 | | 0.4054 | 7.96 | 2950 | 0.7499 | 0.8558 | 0.8692 | 0.8558 | 0.8512 | 0.8980 | | 0.3625 | 8.09 | 3000 | 0.6539 | 0.8639 | 0.8819 | 0.8639 | 0.8606 | 0.9047 | | 0.3625 | 8.23 | 3050 | 0.6250 | 0.8787 | 0.8910 | 0.8787 | 0.8766 | 0.9151 | | 0.3625 | 8.36 | 3100 | 0.6734 | 0.8639 | 0.8709 | 0.8639 | 0.8594 | 0.9053 | | 0.3625 | 8.5 | 3150 | 0.7044 | 0.8571 | 0.8673 | 0.8571 | 0.8519 | 0.9001 | | 0.3625 | 8.63 | 3200 | 0.7766 | 0.8558 | 0.8705 | 0.8558 | 0.8531 | 0.8993 | | 0.3625 | 8.77 | 3250 | 0.6379 | 0.8585 | 0.8723 | 0.8585 | 0.8555 | 0.9001 | | 0.3625 | 8.9 | 3300 | 0.6725 | 0.8612 | 0.8786 | 0.8612 | 0.8577 | 0.9030 | | 0.3322 | 9.04 | 3350 | 0.6457 | 0.8679 | 0.8835 | 0.8679 | 0.8649 | 0.9067 | | 0.3322 | 9.17 | 3400 | 0.7583 | 0.8558 | 0.8671 | 0.8558 | 0.8526 | 0.8982 | | 0.3322 | 9.31 | 3450 | 0.6929 | 0.8598 | 0.8737 | 0.8598 | 0.8538 | 0.9007 | | 0.3322 | 9.44 | 3500 | 0.7696 | 0.8491 | 0.8615 | 0.8491 | 0.8447 | 0.8927 | | 0.3322 | 9.58 | 3550 | 0.7551 | 0.8531 | 0.8725 | 0.8531 | 0.8467 | 0.8973 | | 0.3322 | 9.71 | 3600 | 0.6499 | 0.8598 | 0.8725 | 0.8598 | 0.8565 | 0.9011 | | 0.3322 | 9.84 | 3650 | 0.6816 | 0.8558 | 0.8702 | 0.8558 | 0.8537 | 0.8992 | | 0.3322 | 9.98 | 3700 | 0.6956 | 0.8598 | 0.8739 | 0.8598 | 0.8566 | 0.9011 | ### Framework versions - Transformers 4.38.2 - Pytorch 2.3.0 - Datasets 2.19.1 - Tokenizers 0.15.1