README.md

metadata

license: mit
sdk: streamlit
sdk_version: 1.53.0

XTTS v2 Turkish TTS (Voice Cloning)

Streamlit app running tts_models/multilingual/multi-dataset/xtts_v2 with voice cloning support for Turkish language.

Features

• 🎙️ Text-to-speech synthesis using XTTS v2 model
• 🎭 Voice cloning from uploaded WAV files
• 🇹🇷 Turkish language support
• 📥 Download generated audio files

Running locally

pip install -r requirements.txt
streamlit run app.py

On HuggingFace Spaces

This repo is designed as a Streamlit Space. After creating a new Space and selecting Streamlit, point it to this repository.

Setup Instructions

1. Go to HuggingFace Spaces
2. Click "Create new Space"
3. Select Streamlit as the SDK
4. Connect this repository (or push via the Space's git URL)
5. HuggingFace will automatically install dependencies and run streamlit run app.py

Usage

1. Enter Turkish text in the text area
2. Upload a WAV file containing the speaker's voice (for voice cloning)
3. Click "🔊 Ses üret" to generate speech
4. Listen to the generated audio or download it as a WAV file

Notes

• The first model load may take some time (cold start)
• The model uses CPU by default (gpu=False). For GPU support, change the Space hardware settings and set gpu=True in the code.