stanchino/zonos repository overview

⁠Zonos Text-to-Speech Docker Image

A Docker container for deploying the Zonos-v0.1 open-weight text-to-speech model with CUDA support, Python dependencies managed by uv, and secure SSH access.

⁠Key Features

• 🚀 CUDA-enabled environment for GPU acceleration
• 🐍 Python environment managed by uv package manager
• 🔒 Secure SSH access with public key authentication
• 🌐 Gradio web interface exposed on port 7860
• 🛠️ Built-in development environment with gcc compiler
• 🔄 Weekly automated builds

⁠Quick Start

docker run -it --gpus=all \
  -p 22:22 -p 7860:7860 \
  -e PUBLIC_KEY="$(cat ~/.ssh/id_rsa.pub)" \
  stanchino/zonos:latest

⁠Tags

• latest: Most recent build from main branch
• YYYYMMDD: Weekly builds tagged with date
• YYYYMMDD-[sha]: Commit-triggered builds

⁠Ports

• 7860: Gradio web interface
• 22: SSH access

⁠Environment Variables

• PUBLIC_KEY: Your SSH public key for authentication (required for SSH access)

⁠RunPod Compatible

This image is optimized for RunPod deployment:

• Minimum 6GB VRAM required
• All logs redirected to container stdout
• Accessible via RunPod's proxy URLs

⁠Requirements

• NVIDIA GPU with compatible drivers
• NVIDIA Container Toolkit installed
• Minimum 20GB container disk space
• 6GB+ GPU VRAM

⁠Security

• Password authentication disabled
• Only public key SSH authentication allowed
• Root login restricted to public key authentication

⁠Links

• GitHub Repository⁠
• Zonos Model⁠
• Issue Tracker⁠

⁠License

Apache License 2.0

⁠Support

For issues and feature requests, please use the GitHub issue tracker⁠.