TTS-VoiceMark/Phonon-40M-es/README.md

Phonon-40-es

Just in case:

The safe tensor file (the x6lttij... file) is for the voice of manito. Phonon uses that to cloning the voice and using it for audio and streaming.

Before running.

This model of Phonon-40-es calls huggingface as its a gated model, so you need access it and ask for permition to use it.
This is the link. https://huggingface.co/gradium/phonon-40m-es

Setup

Copy env.example to .env and then add your Hugging Face token as HF_TOKEN. When you run main.py, it loads .env automatically from the same folder; you do not need to set the environment variable manually in the console.

cp env.example .env
uv sync
uv run python main.py