--- license: apache-2.0 base_model: facebook/wav2vec2-base tags: - generated_from_trainer metrics: - accuracy - precision - recall - f1 model-index: - name: wav2vec2-classifier results: [] --- # wav2vec2-classifier This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.6244 - Accuracy: 0.8471 - Precision: 0.8748 - Recall: 0.8471 - F1: 0.8488 - Binary: 0.8939 ## 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: 5e-05 - 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: 10 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | Binary | |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:------:| | No log | 0.17 | 50 | 4.1183 | 0.0437 | 0.0069 | 0.0437 | 0.0092 | 0.3226 | | No log | 0.35 | 100 | 3.7518 | 0.0461 | 0.0070 | 0.0461 | 0.0098 | 0.3272 | | No log | 0.52 | 150 | 3.5442 | 0.0825 | 0.0303 | 0.0825 | 0.0290 | 0.3527 | | No log | 0.69 | 200 | 3.3675 | 0.1068 | 0.0360 | 0.1068 | 0.0473 | 0.3697 | | No log | 0.86 | 250 | 3.2266 | 0.1214 | 0.0494 | 0.1214 | 0.0498 | 0.3784 | | 3.7123 | 1.04 | 300 | 3.0319 | 0.2015 | 0.1672 | 0.2015 | 0.1388 | 0.4388 | | 3.7123 | 1.21 | 350 | 2.8840 | 0.2694 | 0.2341 | 0.2694 | 0.2132 | 0.4871 | | 3.7123 | 1.38 | 400 | 2.6910 | 0.3859 | 0.3375 | 0.3859 | 0.3187 | 0.5694 | | 3.7123 | 1.55 | 450 | 2.5564 | 0.4223 | 0.4081 | 0.4223 | 0.3680 | 0.5934 | | 3.7123 | 1.73 | 500 | 2.4118 | 0.4684 | 0.4887 | 0.4684 | 0.4192 | 0.6265 | | 3.7123 | 1.9 | 550 | 2.2664 | 0.5170 | 0.4814 | 0.5170 | 0.4632 | 0.6612 | | 2.7681 | 2.07 | 600 | 2.1127 | 0.6214 | 0.6252 | 0.6214 | 0.5918 | 0.7350 | | 2.7681 | 2.24 | 650 | 1.9688 | 0.6165 | 0.6309 | 0.6165 | 0.5828 | 0.7308 | | 2.7681 | 2.42 | 700 | 1.8190 | 0.6553 | 0.6452 | 0.6553 | 0.6192 | 0.7580 | | 2.7681 | 2.59 | 750 | 1.7073 | 0.6626 | 0.6724 | 0.6626 | 0.6309 | 0.7631 | | 2.7681 | 2.76 | 800 | 1.6781 | 0.6553 | 0.6746 | 0.6553 | 0.6282 | 0.7573 | | 2.7681 | 2.93 | 850 | 1.5448 | 0.6942 | 0.7092 | 0.6942 | 0.6662 | 0.7852 | | 2.0663 | 3.11 | 900 | 1.4279 | 0.75 | 0.7486 | 0.75 | 0.7261 | 0.8252 | | 2.0663 | 3.28 | 950 | 1.3753 | 0.7403 | 0.7526 | 0.7403 | 0.7185 | 0.8184 | | 2.0663 | 3.45 | 1000 | 1.2878 | 0.7718 | 0.7685 | 0.7718 | 0.7524 | 0.8396 | | 2.0663 | 3.62 | 1050 | 1.2287 | 0.7670 | 0.7899 | 0.7670 | 0.7543 | 0.8354 | | 2.0663 | 3.8 | 1100 | 1.1488 | 0.7791 | 0.8148 | 0.7791 | 0.7687 | 0.8454 | | 2.0663 | 3.97 | 1150 | 1.1336 | 0.7718 | 0.8118 | 0.7718 | 0.7645 | 0.8396 | | 1.5924 | 4.14 | 1200 | 1.0735 | 0.7840 | 0.8185 | 0.7840 | 0.7763 | 0.8488 | | 1.5924 | 4.31 | 1250 | 1.0218 | 0.8058 | 0.8291 | 0.8058 | 0.7960 | 0.8641 | | 1.5924 | 4.49 | 1300 | 0.9783 | 0.7937 | 0.8176 | 0.7937 | 0.7868 | 0.8556 | | 1.5924 | 4.66 | 1350 | 0.9595 | 0.8083 | 0.8247 | 0.8083 | 0.8020 | 0.8667 | | 1.5924 | 4.83 | 1400 | 0.9167 | 0.8155 | 0.8394 | 0.8155 | 0.8072 | 0.8718 | | 1.2937 | 5.0 | 1450 | 0.8874 | 0.8083 | 0.8392 | 0.8083 | 0.8068 | 0.8658 | | 1.2937 | 5.18 | 1500 | 0.8918 | 0.7913 | 0.8305 | 0.7913 | 0.7893 | 0.8541 | | 1.2937 | 5.35 | 1550 | 0.8706 | 0.7937 | 0.8411 | 0.7937 | 0.7948 | 0.8558 | | 1.2937 | 5.52 | 1600 | 0.8037 | 0.8350 | 0.8804 | 0.8350 | 0.8347 | 0.8845 | | 1.2937 | 5.69 | 1650 | 0.8356 | 0.8010 | 0.8518 | 0.8010 | 0.8019 | 0.8626 | | 1.2937 | 5.87 | 1700 | 0.8261 | 0.8131 | 0.8491 | 0.8131 | 0.8159 | 0.8694 | | 1.0749 | 6.04 | 1750 | 0.7995 | 0.8131 | 0.8568 | 0.8131 | 0.8145 | 0.8721 | | 1.0749 | 6.21 | 1800 | 0.7737 | 0.8301 | 0.8674 | 0.8301 | 0.8315 | 0.8830 | | 1.0749 | 6.38 | 1850 | 0.7524 | 0.8228 | 0.8660 | 0.8228 | 0.8236 | 0.8755 | | 1.0749 | 6.56 | 1900 | 0.7203 | 0.8301 | 0.8728 | 0.8301 | 0.8342 | 0.8830 | | 1.0749 | 6.73 | 1950 | 0.7239 | 0.8398 | 0.8751 | 0.8398 | 0.8426 | 0.8881 | | 1.0749 | 6.9 | 2000 | 0.6872 | 0.8422 | 0.8797 | 0.8422 | 0.8454 | 0.8896 | | 0.916 | 7.08 | 2050 | 0.6973 | 0.8398 | 0.8791 | 0.8398 | 0.8431 | 0.8879 | | 0.916 | 7.25 | 2100 | 0.6895 | 0.8350 | 0.8783 | 0.8350 | 0.8403 | 0.8845 | | 0.916 | 7.42 | 2150 | 0.6613 | 0.8495 | 0.8832 | 0.8495 | 0.8510 | 0.8947 | | 0.916 | 7.59 | 2200 | 0.6550 | 0.8325 | 0.8633 | 0.8325 | 0.8348 | 0.8847 | | 0.916 | 7.77 | 2250 | 0.6565 | 0.8422 | 0.8746 | 0.8422 | 0.8434 | 0.8896 | | 0.916 | 7.94 | 2300 | 0.6689 | 0.8350 | 0.8777 | 0.8350 | 0.8381 | 0.8864 | | 0.7992 | 8.11 | 2350 | 0.6816 | 0.8252 | 0.8624 | 0.8252 | 0.8236 | 0.8786 | | 0.7992 | 8.28 | 2400 | 0.6394 | 0.8447 | 0.8784 | 0.8447 | 0.8465 | 0.8932 | | 0.7992 | 8.46 | 2450 | 0.6732 | 0.8252 | 0.8502 | 0.8252 | 0.8219 | 0.8786 | | 0.7992 | 8.63 | 2500 | 0.6593 | 0.8544 | 0.8836 | 0.8544 | 0.8542 | 0.8990 | | 0.7992 | 8.8 | 2550 | 0.6510 | 0.8374 | 0.8690 | 0.8374 | 0.8365 | 0.8871 | | 0.7992 | 8.97 | 2600 | 0.6500 | 0.8398 | 0.8761 | 0.8398 | 0.8417 | 0.8888 | | 0.7376 | 9.15 | 2650 | 0.6393 | 0.8374 | 0.8665 | 0.8374 | 0.8381 | 0.8871 | | 0.7376 | 9.32 | 2700 | 0.6284 | 0.8422 | 0.8716 | 0.8422 | 0.8435 | 0.8905 | | 0.7376 | 9.49 | 2750 | 0.6225 | 0.8471 | 0.8761 | 0.8471 | 0.8483 | 0.8939 | | 0.7376 | 9.66 | 2800 | 0.6219 | 0.8519 | 0.8836 | 0.8519 | 0.8544 | 0.8973 | | 0.7376 | 9.84 | 2850 | 0.6244 | 0.8471 | 0.8748 | 0.8471 | 0.8488 | 0.8939 | ### Framework versions - Transformers 4.38.2 - Pytorch 2.3.0 - Datasets 2.19.1 - Tokenizers 0.15.1