by forkjoin-ai
Open source · 888 downloads · 1 likes
The Qwen3 TTS 12Hz 1.7B Customvoice model is an optimized version of the Qwen3-TTS-12Hz-1.7B-CustomVoice speech synthesis model designed for edge deployment. It generates natural-sounding speech from text with reduced latency thanks to its GGUF format, which is well-suited for constrained environments. Built to run locally on mobile devices or browsers, it provides a high-performance alternative to cloud-based solutions without requiring an internet connection. Key use cases include embedded voice assistants, voiceover generation for mobile apps or embedded tools, and customizable solutions for specific needs. What sets it apart is its optimization for real-time inference on lightweight devices while maintaining high voice quality and flexibility through its customizable model.
Forkjoin.ai conversion of Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice to GGUF format for edge deployment.
./llama-cli -m qwen3-tts-12hz-1.7b-customvoice-gguf.gguf -p "Your prompt here" -n 256
Create a Modelfile:
FROM ./qwen3-tts-12hz-1.7b-customvoice-gguf.gguf
ollama create qwen3-tts-12hz-1.7b-customvoice-gguf -f Modelfile
ollama run qwen3-tts-12hz-1.7b-customvoice-gguf
Forkjoin.ai runs AI models at the edge -- in-browser, on-device, zero cloud cost. These converted models power real-time inference, speech recognition, and natural language capabilities.
All conversions are optimized for edge deployment within browser and mobile memory constraints.
Apache 2.0 (follows upstream model license)