diff --git a/llmage/llmclient.py b/llmage/llmclient.py index 3b310d3..aa8ad08 100644 --- a/llmage/llmclient.py +++ b/llmage/llmclient.py @@ -158,7 +158,7 @@ async def _inference_generator(request, callerid, callerorgid, 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 ') + debug(f'---{params_kw.stream=}, {llm.stream=} ---use uapi_request ') f = partial(uapi_request, request, llm, callerid, callerorgid, params_kw=params_kw) async for d in f(): yield d diff --git a/wwwroot/v1/chat/completions/index.dspy b/wwwroot/v1/chat/completions/index.dspy index b32958a..4816d29 100644 --- a/wwwroot/v1/chat/completions/index.dspy +++ b/wwwroot/v1/chat/completions/index.dspy @@ -21,6 +21,7 @@ if params_kw.off_peak: userid = await get_user() userorgid = await get_userorgid() if userid is None: + debug(f'not params_kw.prompt and not params_kw.messages,{params_kw=}') return openai_403() if not params_kw.prompt and not params_kw.messages: @@ -37,11 +38,13 @@ where a.llmcatelogid=b.id 'model': params_kw.model or 'qwen3-max' }) if len(recs) == 0: + debug(f'{params_kw.model=} not found') return openai_400() params_kw.llmid = recs[0].id f = await checkCustomerBalance(params_kw.llmid, userorgid) if not f: + debug(f'{userid=} balance not enough') return openai_429() return await env.stream_response(request, gen) diff --git a/wwwroot/vidu_inference.dspy b/wwwroot/vidu_inference.dspy index 7fadc9d..2d3d265 100644 --- a/wwwroot/vidu_inference.dspy +++ b/wwwroot/vidu_inference.dspy @@ -10,7 +10,6 @@ userid = await get_user() userorgid = await get_userorgid() if userid is None: return UiError(title='llm inference', message='Please login first') -""" f = await checkCustomerBalance(params_kw.llmid, userorgid) if not f: return { @@ -19,6 +18,5 @@ if not f: "message": "账户余额不够" } } -""" env = DictObject(**globals()) return await inference(request, env=env)