|
---
|
|
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: 1.0124
|
|
- Accuracy: 0.7574
|
|
- Precision: 0.7773
|
|
- Recall: 0.7574
|
|
- F1: 0.7429
|
|
- Binary: 0.8286
|
|
|
|
## 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 | 4.2519 | 0.0566 | 0.0051 | 0.0566 | 0.0089 | 0.3305 |
|
|
| No log | 0.38 | 100 | 3.9645 | 0.0539 | 0.0068 | 0.0539 | 0.0107 | 0.3321 |
|
|
| No log | 0.58 | 150 | 3.7338 | 0.1105 | 0.0576 | 0.1105 | 0.0536 | 0.3717 |
|
|
| No log | 0.77 | 200 | 3.5648 | 0.1617 | 0.0751 | 0.1617 | 0.0846 | 0.4084 |
|
|
| No log | 0.96 | 250 | 3.4031 | 0.2183 | 0.1101 | 0.2183 | 0.1268 | 0.4488 |
|
|
| 3.9542 | 1.15 | 300 | 3.2478 | 0.2102 | 0.1080 | 0.2102 | 0.1281 | 0.4450 |
|
|
| 3.9542 | 1.34 | 350 | 3.0971 | 0.2992 | 0.1963 | 0.2992 | 0.2044 | 0.5081 |
|
|
| 3.9542 | 1.53 | 400 | 2.9647 | 0.2992 | 0.2154 | 0.2992 | 0.2194 | 0.5100 |
|
|
| 3.9542 | 1.73 | 450 | 2.8175 | 0.3774 | 0.3235 | 0.3774 | 0.3129 | 0.5628 |
|
|
| 3.9542 | 1.92 | 500 | 2.6989 | 0.4528 | 0.3689 | 0.4528 | 0.3734 | 0.6129 |
|
|
| 3.1802 | 2.11 | 550 | 2.5825 | 0.4798 | 0.4392 | 0.4798 | 0.4132 | 0.6345 |
|
|
| 3.1802 | 2.3 | 600 | 2.4777 | 0.4960 | 0.4381 | 0.4960 | 0.4257 | 0.6458 |
|
|
| 3.1802 | 2.49 | 650 | 2.3896 | 0.5067 | 0.4835 | 0.5067 | 0.4514 | 0.6534 |
|
|
| 3.1802 | 2.68 | 700 | 2.2850 | 0.5553 | 0.5319 | 0.5553 | 0.5060 | 0.6873 |
|
|
| 3.1802 | 2.88 | 750 | 2.1647 | 0.5660 | 0.5312 | 0.5660 | 0.5136 | 0.6938 |
|
|
| 2.6685 | 3.07 | 800 | 2.0787 | 0.5660 | 0.5682 | 0.5660 | 0.5299 | 0.6949 |
|
|
| 2.6685 | 3.26 | 850 | 1.9976 | 0.5849 | 0.5871 | 0.5849 | 0.5490 | 0.7078 |
|
|
| 2.6685 | 3.45 | 900 | 1.9318 | 0.6119 | 0.5932 | 0.6119 | 0.5737 | 0.7278 |
|
|
| 2.6685 | 3.64 | 950 | 1.8600 | 0.6065 | 0.5880 | 0.6065 | 0.5629 | 0.7240 |
|
|
| 2.6685 | 3.84 | 1000 | 1.7730 | 0.6496 | 0.6337 | 0.6496 | 0.6135 | 0.7550 |
|
|
| 2.2863 | 4.03 | 1050 | 1.7003 | 0.6765 | 0.6607 | 0.6765 | 0.6427 | 0.7720 |
|
|
| 2.2863 | 4.22 | 1100 | 1.6632 | 0.6658 | 0.6876 | 0.6658 | 0.6364 | 0.7655 |
|
|
| 2.2863 | 4.41 | 1150 | 1.6066 | 0.6739 | 0.6552 | 0.6739 | 0.6371 | 0.7712 |
|
|
| 2.2863 | 4.6 | 1200 | 1.5495 | 0.6900 | 0.6841 | 0.6900 | 0.6616 | 0.7825 |
|
|
| 2.2863 | 4.79 | 1250 | 1.5232 | 0.6631 | 0.6333 | 0.6631 | 0.6295 | 0.7644 |
|
|
| 2.2863 | 4.99 | 1300 | 1.4847 | 0.6900 | 0.6985 | 0.6900 | 0.6668 | 0.7825 |
|
|
| 2.0189 | 5.18 | 1350 | 1.4253 | 0.6927 | 0.6710 | 0.6927 | 0.6586 | 0.7833 |
|
|
| 2.0189 | 5.37 | 1400 | 1.4005 | 0.7008 | 0.6947 | 0.7008 | 0.6729 | 0.7900 |
|
|
| 2.0189 | 5.56 | 1450 | 1.3482 | 0.7143 | 0.7217 | 0.7143 | 0.6926 | 0.7984 |
|
|
| 2.0189 | 5.75 | 1500 | 1.3130 | 0.7089 | 0.7150 | 0.7089 | 0.6866 | 0.7957 |
|
|
| 2.0189 | 5.94 | 1550 | 1.2733 | 0.7332 | 0.7519 | 0.7332 | 0.7117 | 0.8127 |
|
|
| 1.8234 | 6.14 | 1600 | 1.2586 | 0.7278 | 0.7585 | 0.7278 | 0.7101 | 0.8089 |
|
|
| 1.8234 | 6.33 | 1650 | 1.2376 | 0.7251 | 0.7509 | 0.7251 | 0.7067 | 0.8059 |
|
|
| 1.8234 | 6.52 | 1700 | 1.2169 | 0.7251 | 0.7454 | 0.7251 | 0.7070 | 0.8059 |
|
|
| 1.8234 | 6.71 | 1750 | 1.2007 | 0.7358 | 0.7476 | 0.7358 | 0.7162 | 0.8135 |
|
|
| 1.8234 | 6.9 | 1800 | 1.1814 | 0.7278 | 0.7410 | 0.7278 | 0.7069 | 0.8078 |
|
|
| 1.6833 | 7.09 | 1850 | 1.1559 | 0.7358 | 0.7627 | 0.7358 | 0.7184 | 0.8135 |
|
|
| 1.6833 | 7.29 | 1900 | 1.1305 | 0.7574 | 0.7876 | 0.7574 | 0.7432 | 0.8286 |
|
|
| 1.6833 | 7.48 | 1950 | 1.1150 | 0.7385 | 0.7671 | 0.7385 | 0.7259 | 0.8173 |
|
|
| 1.6833 | 7.67 | 2000 | 1.1024 | 0.7412 | 0.7925 | 0.7412 | 0.7269 | 0.8173 |
|
|
| 1.6833 | 7.86 | 2050 | 1.0954 | 0.7358 | 0.7272 | 0.7358 | 0.7111 | 0.8135 |
|
|
| 1.5919 | 8.05 | 2100 | 1.0778 | 0.7547 | 0.7598 | 0.7547 | 0.7342 | 0.8267 |
|
|
| 1.5919 | 8.25 | 2150 | 1.0713 | 0.7385 | 0.7462 | 0.7385 | 0.7192 | 0.8164 |
|
|
| 1.5919 | 8.44 | 2200 | 1.0495 | 0.7574 | 0.7719 | 0.7574 | 0.7405 | 0.8286 |
|
|
| 1.5919 | 8.63 | 2250 | 1.0371 | 0.7574 | 0.7677 | 0.7574 | 0.7400 | 0.8296 |
|
|
| 1.5919 | 8.82 | 2300 | 1.0452 | 0.7655 | 0.7876 | 0.7655 | 0.7540 | 0.8353 |
|
|
| 1.5254 | 9.01 | 2350 | 1.0313 | 0.7601 | 0.7781 | 0.7601 | 0.7450 | 0.8315 |
|
|
| 1.5254 | 9.2 | 2400 | 1.0243 | 0.7628 | 0.7862 | 0.7628 | 0.7494 | 0.8334 |
|
|
| 1.5254 | 9.4 | 2450 | 1.0192 | 0.7520 | 0.7697 | 0.7520 | 0.7348 | 0.8259 |
|
|
| 1.5254 | 9.59 | 2500 | 1.0140 | 0.7628 | 0.7859 | 0.7628 | 0.7490 | 0.8323 |
|
|
| 1.5254 | 9.78 | 2550 | 1.0121 | 0.7601 | 0.7774 | 0.7601 | 0.7453 | 0.8305 |
|
|
| 1.5254 | 9.97 | 2600 | 1.0124 | 0.7574 | 0.7773 | 0.7574 | 0.7429 | 0.8286 |
|
|
|
|
|
|
### Framework versions
|
|
|
|
- Transformers 4.38.2
|
|
- Pytorch 2.3.0
|
|
- Datasets 2.19.1
|
|
- Tokenizers 0.15.1
|
|
|
