[Bugfix] Support passing vllm cli args to online serving in vLLM-Omni (#206)

Signed-off-by: Gaohan123 <gaohan19@huawei.com>
Signed-off-by: Gaohan123 <hgaoaf@connect.ust.hk>
This commit is contained in:
Gao Han authored and GitHub committed 2025-12-08 14:59:53 +08:00
1 parent bbb5a72871
commit 45dc189d8d
5 files changed
+131 -36

No files matched your search

+1 -4
View File
@@ -185,10 +185,7 @@ async def build_async_omni_from_stage_config(
async_omni: Optional[EngineClient] = None
try:
if getattr(args, "stage_configs_path", None):
async_omni = AsyncOmni(model=args.model, stage_configs_path=args.stage_configs_path)
else:
async_omni = AsyncOmni(model=args.model)
async_omni = AsyncOmni(model=args.model, cli_args=args)
# # Don't keep the dummy data in memory
# await async_llm.reset_mm_cache()