Distil Wav2vec2 Adult Child Cls 37m

🚀 DistilWav2Vec2 Adult/Child Speech Classifier 37M

DistilWav2Vec2 Adult/Child Speech Classifier is an audio classification model based on the wav2vec 2.0 architecture. It's a distilled version of another model on a private adult/child speech classification dataset.

🚀 Quick Start

This section provides a high - level overview of the model. The DistilWav2Vec2 Adult/Child Speech Classifier is an audio classification model. It is based on the wav2vec 2.0 architecture and is a distilled version of wav2vec2-adult-child-cls on a private adult/child speech classification dataset. The model was trained using HuggingFace's PyTorch framework on a Tesla P100 provided by Kaggle, and training metrics were logged via Tensorboard.

✨ Features

• Based on wav2vec 2.0: Utilizes the advanced wav2vec 2.0 architecture for audio classification.
• Distilled Model: A distilled version of a larger model, potentially more efficient.
• Trained on Private Dataset: Trained on a private adult/child speech classification dataset.

📚 Documentation

Model

Evaluation Results

The model achieves the following results on evaluation:

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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5

Training results

⚠️ Important Note

Do consider the biases which came from pre - training datasets that may be carried over into the results of this model.

📄 License

This model is licensed under the Apache 2.0 license.

👨‍💻 Authors

DistilWav2Vec2 Adult/Child Speech Classifier was trained and evaluated by Ananto Joyoadikusumo. 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