Candle Qwen3-TTS

A Rust implementation of Qwen3-TTS text-to-speech model for the Candle ML framework.

This crate provides:

• High-level model API (model::Model)
• Speaker encoder (ECAPA-TDNN based)
• Audio tokenizer (12Hz)

Architecture Overview

Qwen3-TTS uses a hierarchical generation approach:

1. Text is processed through a talker model with multimodal RoPE
2. The first codebook (semantic) is predicted by the main talker
3. Remaining codebooks (acoustic) are predicted by a sub-talker (code predictor)
4. All 32 codes are decoded to audio via the tokenizer

Example

use qwen_tts::model::loader::{ModelLoader, LoaderConfig};
use candle_core::Device;

# fn main() -> Result<(), Box<dyn std::error::Error>> {
let loader = ModelLoader::from_local_dir("/path/to/model")?;
let model = loader.load_tts_model(&Device::Cpu, &LoaderConfig::default())?;
let result = model.generate_custom_voice_from_text(
    "Hello, world!",
    "vivian",
    "english",
    None,
    None,
)?;
# Ok(())
# }