File size: 6,964 Bytes
f4fecfc c997970 f4fecfc 4e51374 f4fecfc 4e51374 f4fecfc c997970 ec1bd1c f4fecfc c997970 ec1bd1c c997970 f4fecfc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 |
---
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: wav2vec2-classifier-aug
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# wav2vec2-classifier-aug
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: 1.0269
- Accuracy: 0.7790
- Precision: 0.7971
- Recall: 0.7790
- F1: 0.7657
- Binary: 0.8453
## 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: 3e-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 | 0.19 | 50 | 4.2076 | 0.0512 | 0.0054 | 0.0512 | 0.0092 | 0.3245 |
| No log | 0.38 | 100 | 3.9082 | 0.0566 | 0.0053 | 0.0566 | 0.0092 | 0.3364 |
| No log | 0.58 | 150 | 3.7121 | 0.0809 | 0.0281 | 0.0809 | 0.0302 | 0.3553 |
| No log | 0.77 | 200 | 3.5137 | 0.1671 | 0.0794 | 0.1671 | 0.0870 | 0.4164 |
| No log | 0.96 | 250 | 3.3647 | 0.2049 | 0.1259 | 0.2049 | 0.1261 | 0.4429 |
| 3.9294 | 1.15 | 300 | 3.2045 | 0.2534 | 0.1676 | 0.2534 | 0.1635 | 0.4765 |
| 3.9294 | 1.34 | 350 | 3.0682 | 0.2695 | 0.1968 | 0.2695 | 0.1841 | 0.4879 |
| 3.9294 | 1.53 | 400 | 2.9211 | 0.3154 | 0.2301 | 0.3154 | 0.2250 | 0.5191 |
| 3.9294 | 1.73 | 450 | 2.7973 | 0.3720 | 0.2756 | 0.3720 | 0.2815 | 0.5615 |
| 3.9294 | 1.92 | 500 | 2.6856 | 0.4232 | 0.3382 | 0.4232 | 0.3413 | 0.5973 |
| 3.1851 | 2.11 | 550 | 2.5696 | 0.4582 | 0.4161 | 0.4582 | 0.3827 | 0.6208 |
| 3.1851 | 2.3 | 600 | 2.4666 | 0.4987 | 0.4339 | 0.4987 | 0.4215 | 0.6501 |
| 3.1851 | 2.49 | 650 | 2.3548 | 0.5202 | 0.4502 | 0.5202 | 0.4542 | 0.6633 |
| 3.1851 | 2.68 | 700 | 2.2498 | 0.5229 | 0.4593 | 0.5229 | 0.4574 | 0.6660 |
| 3.1851 | 2.88 | 750 | 2.1579 | 0.5660 | 0.5162 | 0.5660 | 0.4993 | 0.6962 |
| 2.7 | 3.07 | 800 | 2.0626 | 0.5903 | 0.5465 | 0.5903 | 0.5332 | 0.7143 |
| 2.7 | 3.26 | 850 | 1.9820 | 0.6092 | 0.5576 | 0.6092 | 0.5470 | 0.7264 |
| 2.7 | 3.45 | 900 | 1.9158 | 0.6011 | 0.5636 | 0.6011 | 0.5466 | 0.7208 |
| 2.7 | 3.64 | 950 | 1.8432 | 0.6092 | 0.5631 | 0.6092 | 0.5499 | 0.7264 |
| 2.7 | 3.84 | 1000 | 1.7732 | 0.6280 | 0.5956 | 0.6280 | 0.5816 | 0.7396 |
| 2.3404 | 4.03 | 1050 | 1.7214 | 0.6523 | 0.6167 | 0.6523 | 0.6000 | 0.7566 |
| 2.3404 | 4.22 | 1100 | 1.6562 | 0.6550 | 0.6312 | 0.6550 | 0.6114 | 0.7585 |
| 2.3404 | 4.41 | 1150 | 1.5909 | 0.6792 | 0.6402 | 0.6792 | 0.6301 | 0.7755 |
| 2.3404 | 4.6 | 1200 | 1.5455 | 0.6900 | 0.6806 | 0.6900 | 0.6534 | 0.7830 |
| 2.3404 | 4.79 | 1250 | 1.5123 | 0.6739 | 0.6415 | 0.6739 | 0.6330 | 0.7717 |
| 2.3404 | 4.99 | 1300 | 1.4662 | 0.7089 | 0.6954 | 0.7089 | 0.6741 | 0.7962 |
| 2.089 | 5.18 | 1350 | 1.4212 | 0.6981 | 0.6739 | 0.6981 | 0.6585 | 0.7887 |
| 2.089 | 5.37 | 1400 | 1.3848 | 0.7008 | 0.6700 | 0.7008 | 0.6572 | 0.7906 |
| 2.089 | 5.56 | 1450 | 1.3435 | 0.7305 | 0.7289 | 0.7305 | 0.7017 | 0.8113 |
| 2.089 | 5.75 | 1500 | 1.3324 | 0.7251 | 0.7331 | 0.7251 | 0.7008 | 0.8075 |
| 2.089 | 5.94 | 1550 | 1.3030 | 0.7116 | 0.7242 | 0.7116 | 0.6841 | 0.7981 |
| 1.8929 | 6.14 | 1600 | 1.2662 | 0.7358 | 0.7356 | 0.7358 | 0.7065 | 0.8151 |
| 1.8929 | 6.33 | 1650 | 1.2341 | 0.7332 | 0.7665 | 0.7332 | 0.7128 | 0.8132 |
| 1.8929 | 6.52 | 1700 | 1.2299 | 0.7224 | 0.7255 | 0.7224 | 0.6950 | 0.8057 |
| 1.8929 | 6.71 | 1750 | 1.1984 | 0.7574 | 0.7758 | 0.7574 | 0.7409 | 0.8302 |
| 1.8929 | 6.9 | 1800 | 1.1810 | 0.7547 | 0.7710 | 0.7547 | 0.7390 | 0.8283 |
| 1.7722 | 7.09 | 1850 | 1.1510 | 0.7763 | 0.7979 | 0.7763 | 0.7613 | 0.8434 |
| 1.7722 | 7.29 | 1900 | 1.1476 | 0.7466 | 0.7541 | 0.7466 | 0.7259 | 0.8226 |
| 1.7722 | 7.48 | 1950 | 1.1326 | 0.7601 | 0.7798 | 0.7601 | 0.7456 | 0.8321 |
| 1.7722 | 7.67 | 2000 | 1.1218 | 0.7655 | 0.7778 | 0.7655 | 0.7493 | 0.8358 |
| 1.7722 | 7.86 | 2050 | 1.0964 | 0.7736 | 0.7844 | 0.7736 | 0.7597 | 0.8415 |
| 1.6621 | 8.05 | 2100 | 1.0895 | 0.7682 | 0.7778 | 0.7682 | 0.7505 | 0.8377 |
| 1.6621 | 8.25 | 2150 | 1.0731 | 0.7682 | 0.7925 | 0.7682 | 0.7546 | 0.8377 |
| 1.6621 | 8.44 | 2200 | 1.0673 | 0.7628 | 0.7771 | 0.7628 | 0.7497 | 0.8340 |
| 1.6621 | 8.63 | 2250 | 1.0728 | 0.7682 | 0.7890 | 0.7682 | 0.7552 | 0.8377 |
| 1.6621 | 8.82 | 2300 | 1.0449 | 0.7898 | 0.8032 | 0.7898 | 0.7779 | 0.8528 |
| 1.6107 | 9.01 | 2350 | 1.0386 | 0.7871 | 0.8037 | 0.7871 | 0.7744 | 0.8509 |
| 1.6107 | 9.2 | 2400 | 1.0398 | 0.7763 | 0.7912 | 0.7763 | 0.7635 | 0.8434 |
| 1.6107 | 9.4 | 2450 | 1.0324 | 0.7844 | 0.8025 | 0.7844 | 0.7726 | 0.8491 |
| 1.6107 | 9.59 | 2500 | 1.0323 | 0.7790 | 0.8006 | 0.7790 | 0.7651 | 0.8453 |
| 1.6107 | 9.78 | 2550 | 1.0274 | 0.7790 | 0.7980 | 0.7790 | 0.7648 | 0.8453 |
| 1.6107 | 9.97 | 2600 | 1.0269 | 0.7790 | 0.7971 | 0.7790 | 0.7657 | 0.8453 |
### Framework versions
- Transformers 4.38.2
- Pytorch 2.3.0
- Datasets 2.19.1
- Tokenizers 0.15.1
|