This is heavily a work in progress.

Completed: Generic Dataset Completed: AN4 Completed: LibriSpeech Completed: Basic feature extraction Completed: Noise injection TODO: Rewrite documentation. TODO: Write more TODOs!

Below is the content of the README.md file from SeanNaren's code, for posterity (for the time being).

This repository consists of:

• torchaudio.data : Generic data loaders for audio
• torchaudio.datasets : Pre-built loaders for common audio Datasets

From source:

.. code:: bash

python setup.py install

The following dataset loaders are available:

• AN4 Audio Database<#an4>
• LibriSpeech <#libri>

AN4

TODO
LibriSpeech
~~~~

TODO