biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Feat][CLI] enforce-include-usage (#19695)

Browse Source 
Signed-off-by: Max Wittig <max.wittig@siemens.com>
This commit is contained in:
Max Wittig
2025-06-25 07:43:04 +02:00
committed by GitHub
parent 879f69bed3
commit f59fc60fb3
5 changed files with 34 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/entrypoints/openai/api_server.py
View File

@@ -1190,6 +1190,7 @@ async def init_app_state(
tool_parser=args.tool_call_parser,
reasoning_parser=args.reasoning_parser,
enable_prompt_tokens_details=args.enable_prompt_tokens_details,
enable_force_include_usage=args.enable_force_include_usage,
) if model_config.runner_type == "generate" else None
state.openai_serving_completion = OpenAIServingCompletion(
engine_client,
@@ -1197,6 +1198,7 @@ async def init_app_state(
state.openai_serving_models,
request_logger=request_logger,
return_tokens_as_token_ids=args.return_tokens_as_token_ids,
enable_force_include_usage=args.enable_force_include_usage,
) if model_config.runner_type == "generate" else None
state.openai_serving_pooling = OpenAIServingPooling(
engine_client,