xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-09 22:44:54 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Misc] Make handling of SamplingParams clearer in n>1 case (#26032)

Browse Source 
Signed-off-by: Nick Hill <nhill@redhat.com>
This commit is contained in:
Nick Hill 2025-10-01 19:31:39 -07:00 committed by GitHub
parent 0b018d8baf
commit 169313b9f8
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 11 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
vllm/v1/engine/async_llm.py
View File

@ -289,13 +289,19 @@ class AsyncLLM(EngineClient):
await self._add_request(request, prompt_str, None, 0, queue)
return queue
# Get the updated SamplingParams from the request, which
# were cloned/updated in processor.process_inputs above.
parent_params = request.sampling_params
assert parent_params is not None
# Fan out child requests (for n>1).
parent_request = ParentRequest(request_id, request.sampling_params)
for idx in range(params.n):
request_id, params = parent_request.get_child_info(idx)
child_request = request if idx == params.n - 1 else copy(request)
parent_request = ParentRequest(request_id, parent_params)
for idx in range(parent_params.n):
request_id, child_params = parent_request.get_child_info(idx)
child_request = request if idx == parent_params.n - 1 else copy(
request)
child_request.request_id = request_id
child_request.sampling_params = params
child_request.sampling_params = child_params
await self._add_request(child_request, prompt_str, parent_request,
idx, queue)
return queue