@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / 248e33c40d
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

[Bugfix][Responses API] Set id on function_call item in streaming done event (#44608)

Signed-off-by: Aniruddh Krovvidi <aniruddh.krovvidi@oracle.com> Co-authored-by: Flora Feng <4florafeng@gmail.com>

ankrovv committed Jun 10, 2026 at 20:52 UTC 248e33c40d3ec3ec605c1dca9419e82d2e733ca6
1 file changed +1 -1
vllm/entrypoints/openai/responses/streaming_events.py
+1 -1
@@ -491,7 +491,7 @@ def emit_function_call_done_events(
491 type="function_call",
492 arguments=arguments,
493 name=function_name,
494 - item_id=state.current_item_id,
494 + id=state.current_item_id,
495 output_index=state.current_output_index,
496 sequence_number=-1,
497 call_id=state.current_call_id,