# Orpheus Music Transformer
## [Announcement] New SOTA multi-instrumental music transformer with advanced ultra-efficient full MIDI encoding, chords tokens and intro/outro tokens
![Orpheus-Music-Transformer-Artwork-1](https://github.com/user-attachments/assets/d358d7d3-ed0a-466a-96ac-58bf1b4ed46a)
***
## Spotlight features
* Advanced ultra-efficient full MIDI encoding (~10% more pitches per 8k seq len compared to the best exsiting encoding schemes)
* Natural and logical tokens arrangement to maximize model generalization and learning abilities
* Chords tokens for chords-conditioned music generation
* Enhanced intro sequence tokens for more stable and precise seeds generation
* Outro tokens for nice outro generation
* Balanced model size and large training corpus
* RoPE
***
### Project Los Angeles
### Tegridy Code 2024
Raw data
{
"_id": null,
"home_page": "https://github.com/asigalov61/orpheusmusictransformer",
"name": "orpheusmusictransformer",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.6",
"maintainer_email": "Alex Lev <alexlev61@proton.me>",
"keywords": "MIDI, music, music ai, music transformer, artificial instelligence, orpheus",
"author": "Alex Lev",
"author_email": "Alex Lev <alexlev61@proton.me>",
"download_url": "https://files.pythonhosted.org/packages/57/16/27e9e2731d7b9529a8007d33392d048dfe3509bf38d422418b53dae44df4/orpheusmusictransformer-0.0.2.tar.gz",
"platform": null,
"description": "# Orpheus Music Transformer\r\n## [Announcement] New SOTA multi-instrumental music transformer with advanced ultra-efficient full MIDI encoding, chords tokens and intro/outro tokens\r\n\r\n![Orpheus-Music-Transformer-Artwork-1](https://github.com/user-attachments/assets/d358d7d3-ed0a-466a-96ac-58bf1b4ed46a)\r\n\r\n***\r\n\r\n## Spotlight features\r\n\r\n* Advanced ultra-efficient full MIDI encoding (~10% more pitches per 8k seq len compared to the best exsiting encoding schemes)\r\n* Natural and logical tokens arrangement to maximize model generalization and learning abilities\r\n* Chords tokens for chords-conditioned music generation\r\n* Enhanced intro sequence tokens for more stable and precise seeds generation\r\n* Outro tokens for nice outro generation\r\n* Balanced model size and large training corpus\r\n* RoPE\r\n \r\n\r\n***\r\n\r\n### Project Los Angeles\r\n### Tegridy Code 2024\r\n",
"bugtrack_url": null,
"license": "Apache Software License 2.0",
"summary": "New SOTA multi-instrumental music transformer with advanced ultra-efficient full MIDI encoding, chords tokens and intro/outro tokens",
"version": "0.0.2",
"project_urls": {
"Demo": "https://huggingface.co/spaces/asigalov61/Orpheus-Music-Transformer",
"Discussions": "https://github.com/asigalov61/orpheusmusictransformer/discussions",
"Documentation": "https://github.com/asigalov61/orpheusmusictransformer/docs",
"Examples": "https://github.com/asigalov61/tegridymidi/orpheusmusictransformer/examples",
"Homepage": "https://github.com/asigalov61/orpheusmusictransformer",
"Issues": "https://github.com/asigalov61/orpheusmusictransformer/issues",
"Models": "https://huggingface.co/asigalov61/Orpheus-Music-Transformer",
"Repository": "https://github.com/asigalov61/Orpheus-Music-Transformer"
},
"split_keywords": [
"midi",
" music",
" music ai",
" music transformer",
" artificial instelligence",
" orpheus"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "8cb0dfbe645f07c00353dd9c892148259d9db2234a7308f614ffcf95735e1cd3",
"md5": "1d37dd1efd0473f171d8613a1c64e411",
"sha256": "364e799e173612ede3e1a6b81df8a38e6ae4db12577cc03f31d1a342791e1f0d"
},
"downloads": -1,
"filename": "orpheusmusictransformer-0.0.2-py3-none-any.whl",
"has_sig": false,
"md5_digest": "1d37dd1efd0473f171d8613a1c64e411",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.6",
"size": 148092,
"upload_time": "2024-12-07T22:10:25",
"upload_time_iso_8601": "2024-12-07T22:10:25.970197Z",
"url": "https://files.pythonhosted.org/packages/8c/b0/dfbe645f07c00353dd9c892148259d9db2234a7308f614ffcf95735e1cd3/orpheusmusictransformer-0.0.2-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "571627e9e2731d7b9529a8007d33392d048dfe3509bf38d422418b53dae44df4",
"md5": "6ae1427dc9f295878bd169ebdcfb3d0d",
"sha256": "4474eb01d33c2f6f022216b052e68f66bf8626dc33549dd08273418cbfe1d9ea"
},
"downloads": -1,
"filename": "orpheusmusictransformer-0.0.2.tar.gz",
"has_sig": false,
"md5_digest": "6ae1427dc9f295878bd169ebdcfb3d0d",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.6",
"size": 144132,
"upload_time": "2024-12-07T22:10:28",
"upload_time_iso_8601": "2024-12-07T22:10:28.114069Z",
"url": "https://files.pythonhosted.org/packages/57/16/27e9e2731d7b9529a8007d33392d048dfe3509bf38d422418b53dae44df4/orpheusmusictransformer-0.0.2.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-12-07 22:10:28",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "asigalov61",
"github_project": "orpheusmusictransformer",
"github_not_found": true,
"lcname": "orpheusmusictransformer"
}