From 51b9c8797654cf8dd0df27c8d9e9211442a464a0 Mon Sep 17 00:00:00 2001 From: yumoqing Date: Sun, 19 Apr 2026 14:06:14 +0800 Subject: [PATCH] bugfix --- llmage/llmclient.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/llmage/llmclient.py b/llmage/llmclient.py index 30b5722..3b310d3 100644 --- a/llmage/llmclient.py +++ b/llmage/llmclient.py @@ -154,9 +154,11 @@ async def _inference_generator(request, callerid, callerorgid, f = partial(async_uapi_request, request, llm, callerid, callerorgid, params_kw=params_kw) elif not params_kw.stream: llm.stream = False + debug(f'---{params_kw.stream=}, {llm.stream=} ---use sync_uapi_request ') f = partial(sync_uapi_request, request, llm, callerid, callerorgid, params_kw=params_kw) else: llm.stream = True + debug(f'---{params_kw.stream=}, {llm.stream=} ---use sync_uapi_request ') f = partial(uapi_request, request, llm, callerid, callerorgid, params_kw=params_kw) async for d in f(): yield d