# iver56/torch-audiomentations

Fast audio data augmentation in PyTorch. Inspired by audiomentations. Useful for deep learning.

Repository: https://github.com/iver56/torch-audiomentations
Canonical: https://ross.abutalabs.com/products/torch-audiomentations
Language: Python
License: MIT
License Family: permissive
Topics: data-augmentation, pytorch, audio-data-augmentation, audio, waveform, dsp, audio-effects, machine-learning, deep-learning, music, sound, sound-processing, augmentation, python, differentiable-data-augmentation
Last push: 2025-11-24T20:07:26+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 53, release rhythm 40, longevity 100
- inputs: {"age_days": 2263, "days_push": 282, "days_rel": 595, "gap_med": 5, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1167, forks 102 (observed 2026-08-28T04:03:50.739889+00:00)

## What it is
A PyTorch library for fast audio data augmentation, inspired by audiomentations. It provides GPU-accelerated, differentiable audio transforms that operate on batches of multichannel audio and can be integrated as nn.Module components in neural networks.

## Use cases
- augment audio training data for deep learning models
- apply gain and polarity inversion to batches of audio tensors on GPU
- speed up audio augmentation for speech recognition training
- use differentiable audio transforms inside a PyTorch model
- augment multichannel stereo audio for music or sound classification
- replace CPU-based audiomentations with a faster GPU pipeline

## When to choose
- you train audio models in PyTorch and need fast, GPU-accelerated augmentation
- you want differentiable audio transforms as part of your model
- you need to augment batches of multichannel audio efficiently
- you already use audiomentations but need better performance

## When to avoid
- you work outside PyTorch (e.g. TensorFlow or NumPy-only pipelines)
- you need multi-GPU/DDP support, which is not officially supported
- you rely on multiprocessing data loading, which can leak memory
- you need stable target-data processing, which is still experimental

## Facets
- artifact type: library
- maturity: active
- function: audio-processing, machine-learning, deep-learning, data-generation
- domain: machine-learning, deep-learning, speech-processing
- platform: python, cross-platform
- tags: pytorch, data-augmentation, audio-effects, dsp, waveform, differentiable, gpu-accelerated, batch-processing, audio, gpu

## Member repositories
- iver56/torch-audiomentations (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:50.739889+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:28:51.684536+00:00, confidence not recorded.
  - readme: https://github.com/iver56/torch-audiomentations (fetched 2026-08-28T04:03:50.739889+00:00, sha 1f38dcf22fce)
- Data as of 2026-08-30T08:39:29.467469+00:00.