Inference Cookbook¶

Choose a complete recipe maintained in the FastVideo repository. Each command runs its checked-in source directly, so coupled model, GPU, offload, and attention settings do not drift into unsupported combinations.

The commands expect a local clone:

git clone https://github.com/hao-ai-lab/FastVideo.git
cd FastVideo

Loading…

Customize a recipe¶

Start from the checked-in source, then change only the settings your model supports:

• Configuration covers the Python and CLI config surfaces.
• Optimizations covers attention backends, compilation, and memory tradeoffs.
• Support matrix lists supported models and optimizations.