Expand description
Per-request dispatch for dense, sparse, dual, probe, and adaptive warmup.
Structsยง
- Dispatch
Outcome ๐ - Result of dispatching one worker request through inference.
Functionsยง
- dispatch_
request ๐ - Runs one
EmbedRequestagainst a loaded session. - reply_
request_ ๐load_ error - Sends a model-reload error to whichever reply channel the request carries.