LibreAI / Música com IA / HS-TasNet
// ai-music · lucidrains/HS-TasNet

HS-TasNet

HS-TasNet é um projeto de música com IA de código aberto de lucidrains (lucidrains/HS-TasNet no GitHub). Tem 112 estrelas no GitHub e 15 forks, é publicado sob a licença MIT e foi atualizado pela última vez no GitHub em 25 de agosto de 2026.

Real-time, low-latency music source separation in PyTorch.

💎 Joia escondida★ 112 estrelas⑂ 15 forksMITatualizado ago 2026
▦ visual: o README deste projeto não inclui nenhuma captura incorporável,
então aqui vai um marcador. O guia abaixo está completo mesmo assim.
// dados do projeto

Dados do projeto

Estrelas no GitHub
112
Forks
15
Licença
MIT
Última atualização no GitHub
25 de agosto de 2026
Tipo de ficha
Joia escondida: uma página mais breve para um projeto menos conhecido
Site oficial
GitHub
// como funciona

Como funciona

  1. Clone the repo

    Get the implementation.

  2. Install PyTorch deps

    Set up the training environment.

  3. Prepare stems data

    Get separated-track training data.

  4. Train or infer

    Run separation on your mixes.

// recursos

Por que as pessoas gostam tanto

Source separation

Splits mixes into stems.

Real-time design

Low-latency architecture.

Paper implementation

Hybrid Spectrogram-TasNet in code.

PyTorch

Standard deep-learning stack.

Music focused

Tuned for musical sources.

MIT licensed

Free for research and products.

Advertisement
// execute você mesmo

Execute você mesmo

  1. Step 1 — Clone the repo

    Download the source and enter the folder.

    git clone https://github.com/lucidrains/HS-TasNet.git
    cd HS-TasNet
  2. Step 2 — Install deps

    Set up the Python environment.

    pip install -r requirements.txt
  3. Step 3 — Run the model

    Follow the README training/inference commands.

    # see README for the exact training command
// casos de uso

O que você pode criar

  • Separate vocals from a music mix.
  • Build real-time stem-splitting tools.
  • Research low-latency separation.
  • Prototype karaoke-style vocal removal.
Advertisement
// perguntas frequentes

Perguntas frequentes

What is HS-TasNet?

A PyTorch implementation of real-time music source separation, at lucidrains/HS-TasNet.

What does source separation mean?

Splitting a mixed track into vocals, drums, bass, and so on.

Is it really real-time?

The architecture targets low latency; actual speed depends on hardware.

Is it free?

Yes. MIT licensed.

Do I need training data?

Yes. Separated stems are needed to train; see the README.

// continue explorando

Ferramentas semelhantes

Música com IA

instruct-MusicGen

Research code for instruction-based text-to-music editing of music models.

★ 111 estrelasApache-2.0
Ver o guia →
Música com IA

minimaxmusic.cpp

Run the MiniMax Music 3 generator locally in C++.

★ 35 estrelasMIT
Ver o guia →
Áudio

pyannote-audio

Neural building blocks for speaker diarization: speech activity detection, speaker change detection, overlapped speech detection, speaker embedding

★ 10,6 mil estrelasMIT
Ver o guia →
Música com IA

stemgen-rt

Run stemgen-rt locally: stemgenrt is a real-time low-latency music source separ.

★ 46 estrelasMIT
Ver o guia →
Música com IA

moises-light

A resource-efficient music source separation model in PyTorch.

★ 36 estrelasMIT
Ver o guia →
Música com IA

MSST-WebUI

A WebUI app for Music-Source-Separation-Training and we bundled UVR together!

★ 1,3 mil estrelasAGPL-3.0
Ver o guia →
// alternativas

Alternativas em Música com IA

Os outros projetos de Música com IA com mais estrelas no catálogo do LibreAI — a lista realista de alternativas se HS-TasNet não for o ideal para você.

  • YuE★ 10,8 mil

    YuE2: frontier music generation with symbolic planning, zero-shot covers, and agentic music editing.

  • ace-step-ui★ 5,0 mil

    🎵 The Ultimate Open Source Suno Alternative - Professional UI for ACE-Step 1.5 AI Music Generation. Free, local, unlimited. Stop paying for Suno!

  • ACE-Step★ 5,0 mil

    ACE-Step — the open foundation model for music generation: full songs with vocals and lyrics in 19 languages, up to 4 minutes in about 20 seconds, Apache-2.0

  • DiffSinger★ 4,9 mil

    Official DiffSinger code: diffusion-based singing voice synthesis and text-to-speech (AAAI 2022)

  • Stable Audio Tools★ 3,9 mil

    Stable Audio Tools — training and inference code for Stability's audio generation models, including Stable Audio Open, MIT

// mais em ai-music

Mais em Música com IA

Projetos de Música com IA próximos no catálogo, os mais próximos em estrelas de HS-TasNet: continue explorando dos dois lados dele.

  • coding-with-beat★ 114

    A retro pixel DJ for Claude Code / Codex CLI / Terminal with Apple Music / QQ Music

  • instruct-MusicGen★ 111

    Research code for instruction-based text-to-music editing of music models.

  • AIEV★ 126

    Automatic AI video editing.

  • motion-video-skill★ 110

    Agent skill that produces beat-synced 1080p motion-graphic videos in HyperFrames.

  • ComfyUI_FL-HeartMuLa★ 133

    ComfyUI nodes for multilingual AI songs with lyrics via HeartMuLa.

  • ai-video-workflow★ 106

    One-click AIGC pipeline: image, video, music, and prompts in a single flow.

Baixe o HS-TasNet

Grátis e aberto: pegue o código-fonte, dê uma estrela no repositório e comece a criar.