# bytedance/piano_transcription

Repository: https://github.com/bytedance/piano_transcription
Canonical: https://ross.abutalabs.com/products/piano_transcription
Language: Python
License Family: other
Topics: research
Archived: true
Last push: 2023-08-18T16:41:03+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2198, "days_push": 1111, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2021, forks 259 (observed 2026-08-28T04:06:06.366396+00:00)

## What it is
A PyTorch implementation of ByteDance's high-resolution piano transcription system that converts piano audio recordings into MIDI files. It includes a pretrained model for inference and training code based on the MAESTRO dataset.

## Use cases
- convert piano recordings to MIDI files
- transcribe piano audio into note events
- train a piano transcription model from scratch
- extract MIDI from piano performance audio
- run audio-to-MIDI transcription on GPU

## When to choose
- you need to transcribe solo piano recordings into MIDI
- you want a pretrained high-resolution transcription model
- you want to reproduce or extend research on piano transcription

## When to avoid
- you need transcription of non-piano instruments or polyphonic general music
- you need an actively maintained project with recent updates
- you need a production service without any ML setup

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, audio-processing, deep-learning
- domain: machine-learning
- platform: python, cli
- tags: pytorch, piano-transcription, audio-to-midi, pretrained-model, music-transcription, audio, music-information-retrieval, gpu

## Member repositories
- bytedance/piano_transcription (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:06.366396+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 2026-08-30T03:00:10.045825+00:00, confidence not recorded.
  - readme: https://github.com/bytedance/piano_transcription (fetched 2026-08-28T04:06:06.366396+00:00, sha eaae15dc1365)
- Data as of 2026-08-30T08:39:29.467469+00:00.