matchboxnet documentation

MatchboxNet – an end-to-end neural network for speech command recognition.

MatchboxNet is a deep residual network composed from blocks of 1D time-channel separable convolution, batch-normalization, ReLU, and dropout layers.

This package is created to provide a ready-to-use implementation of MatchboxNet for audio classification tasks using PyTorch and Hugging Face Transformers.

Contents:

• Installation
• Configuration
• Feature Extractor
• Model For Audio Classification
• Model
• Utils
• Dataset

Indices and tables

• Index

• Search Page