Hugging Face
Deploy Hugging Face models
Paste a repo and see a memory estimate.
NVIDIA CUDA, AMD, and Apple silicon.
The model catalog has a box for a repo ID or URL.
It shows the license, a suggested engine, and the estimate.
You can then propose the model. Proposing does not publish it.
Questions
Where do I paste a repo?
Use the box at the bottom of the model catalog. You see the estimate first.
What about gated repos?
Gated repos need you to sign in with Hugging Face.