Install on Ubuntu 24.04 with CUDA Support

[pluberd]

Chatterbox-TTS – Manual Installation

⚠️ The standard installation may not work in all environments. This guide provides a working manual setup. Chinese language support may require additional verification.

1. Create a Virtual Environment

python -m venv venv
source venv/bin/activate  # Linux / macOS
venv\Scripts\activate     # Windows

2. Install PyTorch

Install your preferred version. Example for CUDA 12.9:

pip install torch torchaudio torchvision --index-url https://download.pytorch.org/whl/cu129

3. Prepare pyproject.toml

[project]
name = "chatterbox-tts"
version = "0.1.5"
requires-python = ">=3.10"
dependencies = [
  "numpy>=2.0",
  "librosa>=0.10.0",
  "transformers>=4.40",
  "diffusers>=0.29.0",
  "resemble-perth>=1.0.1",
  "omegaconf>=2.3.0",
  "conformer>=0.3.2",
]

[project.optional-dependencies]
zh = ["spacy-pkuseg>=1.0.0"]
ja = ["pykakasi>=2.3.0"]

[project.urls]
Homepage = "https://github.com/resemble-ai/chatterbox"

[build-system]
requires = ["setuptools>=68", "wheel", "pip"]
build-backend = "setuptools.build_meta"

⚠️ Torch is not enforced in dependencies – install CPU/CUDA version yourself.

4. Install Project

pip install -e .

5. Install Additional Dependencies

pip install gradio s3tokenizer

✅ Everything should now be functional.

Start: python gradio_tts_app.py

[ismailnamli33]

Thanks :)