Wav2Vec2 Adult/Child Speech Classifier
Wav2Vec2 Adult/Child Speech Classifier is an audio classification model based on the wav2vec 2.0 architecture. This model is a fine-tuned version of wav2vec2-base on a private adult/child speech classification dataset.
This model was trained using HuggingFace's PyTorch framework. All training was done on a Tesla P100, provided by Kaggle. Training metrics were logged via Tensorboard.
Model
Model |
#params |
Arch. |
Training/Validation data (text) |
wav2vec2-adult-child-cls |
91M |
wav2vec 2.0 |
Adult/Child Speech Classification Dataset |
Evaluation Results
The model achieves the following results on evaluation:
Dataset |
Loss |
Accuracy |
F1 |
Adult/Child Speech Classification |
0.1682 |
95.80% |
0.9618 |
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
optimizer
: Adam with betas=(0.9,0.999)
and epsilon=1e-08
lr_scheduler_type
: linear
lr_scheduler_warmup_ratio
: 0.1
num_epochs
: 5
Training results
Training Loss |
Epoch |
Step |
Validation Loss |
Accuracy |
F1 |
0.2709 |
1.0 |
384 |
0.2616 |
0.9104 |
0.9142 |
0.2112 |
2.0 |
768 |
0.1826 |
0.9386 |
0.9421 |
0.1755 |
3.0 |
1152 |
0.1898 |
0.9354 |
0.9428 |
0.0915 |
4.0 |
1536 |
0.1682 |
0.9580 |
0.9618 |
0.1042 |
5.0 |
1920 |
0.1717 |
0.9511 |
0.9554 |
Disclaimer
Do consider the biases which came from pre-training datasets that may be carried over into the results of this model.
Authors
Wav2Vec2 Adult/Child Speech Classifier was trained and evaluated by Wilson Wongso. All computation and development are done on Kaggle.
Framework versions
- Transformers 4.16.2
- Pytorch 1.10.2+cu102
- Datasets 1.18.3
- Tokenizers 0.10.3