# sooftware/conformer

[Unofficial] PyTorch implementation of "Conformer: Convolution-augmented Transformer for Speech Recognition" (INTERSPEECH 2020)

Repository: https://github.com/sooftware/conformer
Canonical: https://ross.abutalabs.com/products/conformer
Language: Python
License: Apache-2.0
License Family: permissive
Topics: conformer, transformer, cnn, transformer-xl, asr, speech-recognition, pytorch, conv, convolution, augmented, speech, recognition
Last push: 2026-06-29T16:05:01+00:00

## Health v2 (maintenance only)
Score: 72/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 90, release rhythm 32, longevity 100
- inputs: {"age_days": 2050, "days_push": 65, "days_rel": 240, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1131, forks 192 (observed 2026-08-28T04:03:42.560632+00:00)

## What it is
An unofficial PyTorch implementation of the Conformer architecture (convolution-augmented Transformer) from the INTERSPEECH 2020 paper, targeting speech recognition. It provides model code only; training is delegated to the openspeech project.

## Use cases
- implement conformer speech recognition model in pytorch
- build an asr encoder with convolution-augmented transformer
- use conformer blocks for audio sequence modeling
- experiment with state-of-the-art speech recognition architectures
- integrate a conformer encoder into my own speech model

## When to choose
- you need the Conformer architecture in PyTorch without training infrastructure
- you want a lightweight, model-only library to plug into your own ASR pipeline
- you are reproducing or extending the INTERSPEECH 2020 Conformer paper

## When to avoid
- you need a full end-to-end ASR training pipeline with datasets and checkpoints
- you want inference-ready pretrained speech recognition models
- you need non-PyTorch frameworks like TensorFlow or JAX

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, speech-recognition
- domain: speech-processing, machine-learning, deep-learning
- platform: python
- tags: pytorch, conformer, asr, transformer, speech-recognition, model-implementation, gpu

## Member repositories
- sooftware/conformer (main) score 72

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:42.560632+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 2026-08-30T06:38:04.401672+00:00, confidence not recorded.
  - readme: https://github.com/sooftware/conformer (fetched 2026-08-28T04:03:42.560632+00:00, sha a2a2315cbe8c)
- Data as of 2026-08-30T08:39:29.467469+00:00.