diff --git a/litellm/chat_completions/dispatch.py b/litellm/chat_completions/dispatch.py index 968cdb5b720..37e188bc0c5 100644 --- a/litellm/chat_completions/dispatch.py +++ b/litellm/chat_completions/dispatch.py @@ -70,17 +70,26 @@ def _public_request( ) +def _context(request: LiteLLMChatCompletionsRequest) -> Context: + return Context( + Route.CHAT_COMPLETIONS, + provider=request.custom_llm_provider, + model=request.model, + delivery=Delivery.STREAMING if request.stream else Delivery.COMPLETED, + ) + + _DISPATCH: Final = PublicDispatch( route=Route.CHAT_COMPLETIONS, request=lambda args, kwargs: _public_request(_COMPLETION, args, kwargs), - context=lambda request: _context(request), + context=_context, bypass=lambda request: request.kwargs.get("acompletion") is True, ) _ADISPATCH: Final = PublicDispatch( route=Route.CHAT_COMPLETIONS, request=lambda args, kwargs: _public_request(_ACOMPLETION, args, kwargs), - context=lambda request: _context(request), + context=_context, ) @@ -109,15 +118,6 @@ async def acompletion(*args: object, **kwargs: object) -> ChatResult: # kwargs- ) -def _context(request: LiteLLMChatCompletionsRequest) -> Context: - return Context( - Route.CHAT_COMPLETIONS, - provider=request.custom_llm_provider, - model=request.model, - delivery=Delivery.STREAMING if request.stream else Delivery.COMPLETED, - ) - - completion.__doc__ = _PYTHON_COMPLETION.__doc__ completion.__wrapped__ = _PYTHON_COMPLETION # pyright: ignore[reportFunctionMemberAccess] # inspect.signature follows __wrapped__ to the legacy signature acompletion.__doc__ = _PYTHON_ACOMPLETION.__doc__ diff --git a/litellm/messages/dispatch.py b/litellm/messages/dispatch.py index c5c5c36593e..61fb869ba8e 100644 --- a/litellm/messages/dispatch.py +++ b/litellm/messages/dispatch.py @@ -69,17 +69,26 @@ def _public_request( ) +def _context(request: LiteLLMMessagesRequest) -> Context: + return Context( + Route.MESSAGES, + provider=request.custom_llm_provider, + model=request.model, + delivery=Delivery.STREAMING if request.stream else Delivery.COMPLETED, + ) + + _DISPATCH: Final = PublicDispatch( route=Route.MESSAGES, request=lambda args, kwargs: _public_request(_MESSAGES, args, kwargs), - context=lambda request: _context(request), + context=_context, bypass=lambda request: request.kwargs.get("is_async") is True, ) _ADISPATCH: Final = PublicDispatch( route=Route.MESSAGES, request=lambda args, kwargs: _public_request(_AMESSAGES, args, kwargs), - context=lambda request: _context(request), + context=_context, ) @@ -108,15 +117,6 @@ async def anthropic_messages(*args: object, **kwargs: object) -> MessagesResult: ) -def _context(request: LiteLLMMessagesRequest) -> Context: - return Context( - Route.MESSAGES, - provider=request.custom_llm_provider, - model=request.model, - delivery=Delivery.STREAMING if request.stream else Delivery.COMPLETED, - ) - - anthropic_messages_handler.__doc__ = _PYTHON_MESSAGES.__doc__ anthropic_messages_handler.__wrapped__ = _PYTHON_MESSAGES # pyright: ignore[reportFunctionMemberAccess] # inspect.signature follows __wrapped__ to the legacy signature anthropic_messages.__doc__ = _PYTHON_AMESSAGES.__doc__ diff --git a/litellm/ocr/dispatch.py b/litellm/ocr/dispatch.py index 3b43eecf001..f5690a917f4 100644 --- a/litellm/ocr/dispatch.py +++ b/litellm/ocr/dispatch.py @@ -49,17 +49,22 @@ _PYTHON_AOCR: Final = cast( # cast-ok: forward the original call shape through Callable[..., Awaitable[OCRResponse]], main.aocr ) + +def _context(request: LiteLLMOcrRequest) -> Context: + return Context(Route.OCR, provider=request.custom_llm_provider, model=request.model) + + _DISPATCH: Final = PublicDispatch( route=Route.OCR, request=lambda args, kwargs: _public_request("ocr", args, kwargs), - context=lambda request: _context(request), + context=_context, bypass=lambda request: request.kwargs.get("aocr") is True, ) _ADISPATCH: Final = PublicDispatch( route=Route.OCR, request=lambda args, kwargs: _public_request("aocr", args, kwargs), - context=lambda request: _context(request), + context=_context, ) @@ -84,7 +89,3 @@ async def aocr(*args: object, **kwargs: object) -> OCRResponse: # kwargs-ok: pr binding=NATIVE_AOCR, native=lambda hook, request, call_args, call_kwargs: hook(request, call_args, call_kwargs), ) - - -def _context(request: LiteLLMOcrRequest) -> Context: - return Context(Route.OCR, provider=request.custom_llm_provider, model=request.model) diff --git a/litellm/responses/dispatch.py b/litellm/responses/dispatch.py index 8c629f9d1f2..f7418df9886 100644 --- a/litellm/responses/dispatch.py +++ b/litellm/responses/dispatch.py @@ -62,17 +62,26 @@ def _public_request( ) +def _context(request: LiteLLMResponsesRequest) -> Context: + return Context( + Route.RESPONSES, + provider=request.custom_llm_provider, + model=request.model, + delivery=Delivery.STREAMING if request.stream else Delivery.COMPLETED, + ) + + _DISPATCH: Final = PublicDispatch( route=Route.RESPONSES, request=lambda args, kwargs: _public_request(_RESPONSES, args, kwargs), - context=lambda request: _context(request), + context=_context, bypass=lambda request: request.kwargs.get("aresponses") is True, ) _ADISPATCH: Final = PublicDispatch( route=Route.RESPONSES, request=lambda args, kwargs: _public_request(_ARESPONSES, args, kwargs), - context=lambda request: _context(request), + context=_context, ) @@ -101,15 +110,6 @@ async def aresponses(*args: object, **kwargs: object) -> ResponsesResult: # kwa ) -def _context(request: LiteLLMResponsesRequest) -> Context: - return Context( - Route.RESPONSES, - provider=request.custom_llm_provider, - model=request.model, - delivery=Delivery.STREAMING if request.stream else Delivery.COMPLETED, - ) - - responses.__doc__ = _PYTHON_RESPONSES.__doc__ responses.__wrapped__ = _PYTHON_RESPONSES # pyright: ignore[reportFunctionMemberAccess] # inspect.signature follows __wrapped__ to the legacy signature aresponses.__doc__ = _PYTHON_ARESPONSES.__doc__