Neural Style Transfer Toolkit
This site is the lightweight entry point for the PyTorch project. Heavy inference (VGG-19, AdaIN, optical flow, Streamlit UI) runs locally or in Docker — Vercel hosts documentation and links here, not the GPU stack.
What ships in the repo
Gatys optimization NST, feed-forward AdaIN, ReCoNet-style video nets, CLI scripts, Streamlit, Docker, and GitHub Actions CI.
Run the full toolkit
Clone the repo, install Python deps (pip install -e .), use transform_image.py / transform_video.py, or docker build + streamlit run streamlit_app.py.