# bytedance/GiantMIDI-Piano

Repository: https://github.com/bytedance/GiantMIDI-Piano
Canonical: https://ross.abutalabs.com/products/giantmidi-piano
Language: Python
License Family: other
Topics: research
Archived: true
Last push: 2022-04-17T19:47:19+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2177, "days_push": 1599, "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 1895, forks 195 (observed 2026-08-28T04:05:50.816748+00:00)

## What it is
GiantMIDI-Piano is a classical piano MIDI dataset containing 10,855 transcribed MIDI files from 2,786 composers, generated by transcribing live YouTube recordings with a high-resolution piano transcription system. The repository also includes scripts to download the source audio and reproduce the transcription pipeline with PyTorch.

## Use cases
- download classical piano MIDI files for music research
- transcribe piano audio recordings into MIDI
- train music transcription models on piano data
- build a dataset of classical piano performances
- analyze MIDI files from thousands of composers
- reproduce the GiantMIDI-Piano transcription pipeline

## When to choose
- you need a large-scale classical piano MIDI corpus for research
- you want to transcribe solo piano audio into MIDI with a high-resolution system
- you need composer-labeled MIDI data for music information retrieval

## When to avoid
- you need MIDI for non-piano or non-classical music
- you cannot agree to the disclaimer or handle YouTube-derived audio licensing constraints
- you lack GPU resources for transcribing audio yourself and cannot use the pre-built download

## Facets
- artifact type: dataset
- maturity: stable
- function: machine-learning, audio-processing, data-generation
- domain: machine-learning, audio
- platform: python, cross-platform
- tags: midi, piano-transcription, classical-music, music-information-retrieval, research-dataset, audio, datasets, gpu

## Member repositories
- bytedance/GiantMIDI-Piano (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:50.816748+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 2026-08-30T03:12:51.294462+00:00, confidence not recorded.
  - readme: https://github.com/bytedance/GiantMIDI-Piano (fetched 2026-08-28T04:05:50.816748+00:00, sha 6521b92df238)
- Data as of 2026-08-30T08:39:29.467469+00:00.