MIDI-ME mirror of MuScriptor β€” Medium

This is an unchanged, noncommercial mirror of MuScriptor Medium. It exists solely to let the free MIDI-ME audio-to-MIDI plugin download a local transcription model on first use without requiring end users to create a Hugging Face account.

Attribution, conditions, and licence

  • Original model: MuScriptor, developed by Mirelo x Kyutai.
  • Original authors: Alexandre Rouard, Michael Krause, Axel Roebel, Carl-Johann Simon-Gabriel, and Alexandre Defossez.
  • Original checkpoint: MuScriptor/muscriptor-medium at f32236969308476e01fd3aae67357de5feb05a2d.
  • Checkpoint licence: Creative Commons Attribution-NonCommercial 4.0 International.

The upstream model card also requires users to have all necessary rights, including intellectual-property rights, to audio they transcribe, and not to use MuScriptor or its output for illegal or unauthorized activity. The full supplemental conditions, disclaimers, intended-use information, limitations, and citation are preserved verbatim in UPSTREAM_MODEL_CARD.md.

LUSKII-VSTS/MIDI-ME did not create, train, own, sponsor, or endorse MuScriptor. The MuScriptor authors do not endorse or affiliate with LUSKII-VSTS or MIDI-ME. The checkpoint and configuration files are unchanged from the cited upstream revision. This mirror grants no rights beyond the original licence and terms.

For a takedown or correction request, open a discussion on this repository; LUSKII-VSTS will review it promptly.

File integrity

  • model.safetensors: 1,228,144,472 bytes
  • SHA-256: ac80adbdf85d87231735fd948af7013441c0afced316c4e9067fd5d8a7fb97ec

Included materials

  • UPSTREAM_MODEL_CARD.md β€” original model card, preserved verbatim.
  • NOTICE.md β€” attribution and mirror notice.
  • LICENSE β€” link to the original CC BY-NC 4.0 legal code.
Downloads last month
-
Safetensors
Model size
0.3B params
Tensor type
F32
Β·
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support