mirror of
https://github.com/BerriAI/litellm.git
synced 2026-10-04 02:31:27 +00:00
enrichPolicyTemplateStream had no AbortSignal parameter, so its while(true) reader loop could not be cancelled: closing the AI suggestion or template parameter modal mid-stream left the SSE connection draining until the server finished, with every event callback still calling setState on the unmounted modal. Thread an optional signal through the helper's options and abort from both callers when the modal is closed, unmounted, or a new enrichment run supersedes an in-flight one. Abort rejections are not reported as errors |
||
|---|---|---|
| .. | ||
| litellm-dashboard | ||
| Dockerfile | ||
| nginx.conf | ||