fix more typing stuff

weave/integrations/langchain/langchain_nvidia_ai_endpoints.py

@@ -5,8 +5,7 @@
 from typing import TYPE_CHECKING, Any, Callable, Optional
 
 if TYPE_CHECKING:
-    from langchain_core.messages import AIMessageChunk, convert_to_openai_messages
-    from openai.types.chat import ChatCompletion
+    pass
 
 import weave
 from weave.trace.op import Op, ProcessedInputs
@@ -16,6 +15,7 @@
 
 # NVIDIA-specific accumulator for parsing the objects of streaming interactions
 def nvidia_accumulator(acc: Optional[Any], value: Any) -> Any:
+    from langchain_core.messages import AIMessageChunk
     from langchain_core.outputs import ChatGenerationChunk
 
     if acc is None:
@@ -31,7 +31,9 @@ def nvidia_accumulator(acc: Optional[Any], value: Any) -> Any:
 
 # Post processor to transform output into OpenAI's ChatCompletion format -- need to handle stream and non-stream outputs
 def post_process_to_openai_format(output: Any) -> dict:
+    from langchain_core.messages import convert_to_openai_messages
     from langchain_core.outputs import ChatGenerationChunk, ChatResult
+    from openai.types.chat import ChatCompletion
 
     if isinstance(output, ChatResult):  ## its ChatResult
         message = output.llm_output  ## List of ChatGeneration
@@ -112,6 +114,8 @@ def post_process_to_openai_format(output: Any) -> dict:
 def process_inputs_to_openai_format_stream(
     func: Op, args: tuple, kwargs: dict
 ) -> ProcessedInputs | None:
+    from langchain_core.messages import convert_to_openai_messages
+
     original_args = args
     original_kwargs = kwargs
 
@@ -143,6 +147,8 @@ def process_inputs_to_openai_format_stream(
 def process_inputs_to_openai_format(
     func: Op, args: tuple, kwargs: dict
 ) -> ProcessedInputs | None:
+    from langchain_core.messages import convert_to_openai_messages
+
     original_args = args
     original_kwargs = kwargs