Skip to content

Glossary: Running Open Models Locally

Slots

Also written: -np, n_slots

The number of requests llama-server can work on at once, set with `-np`. The context is shared between slots, so give each enough.

Where the book teaches it

Related terms

llama serveContext length