Update private_gpt/server/chat/chat_router.py

private_gpt/server/chat/chat_router.py

@@ -2,6 +2,7 @@ from fastapi import APIRouter, Depends, Request
 from llama_index.llms import ChatMessage, MessageRole
 from pydantic import BaseModel
 from starlette.responses import StreamingResponse
+import logging
 
 from private_gpt.open_ai.extensions.context_filter import ContextFilter
 from private_gpt.open_ai.openai_models import (
@@ -107,6 +108,8 @@ class ChatBody(BaseModel):
 #             completion.response, completion.sources if body.include_sources else None
 #         )
 
+logger = logging.getLogger(__name__)
+
 @chat_router.post(
     "/chat/completions",
     response_model=None,
@@ -118,16 +121,19 @@ def chat_completion(
 ) -> OpenAICompletion | StreamingResponse:
     """Given a list of messages comprising a conversation, return a response."""
     try:
+        logger.info("Received chat completion request with body: %s", body.json())
         service = request.state.injector.get(ChatService)
         all_messages = [
             ChatMessage(content=m.content, role=MessageRole(m.role)) for m in body.messages
         ]
+        logger.info("Constructed all_messages: %s", all_messages)
         if body.stream:
             completion_gen = service.stream_chat(
                 messages=all_messages,
                 use_context=body.use_context,
                 context_filter=body.context_filter,
             )
+            logger.info("Streaming response initialized")
             return StreamingResponse(
                 to_openai_sse_stream(
                     completion_gen.response,
@@ -141,11 +147,11 @@ def chat_completion(
                 use_context=body.use_context,
                 context_filter=body.context_filter,
             )
+            logger.info("Completed chat request: %s", completion.response)
             return to_openai_response(
                 completion.response, completion.sources if body.include_sources else None
             )
     except Exception as e:
-        # Log the exception details for debugging
-        print(f"Error processing chat completion: {e}")
+        logger.error("Error processing chat completion: %s", str(e), exc_info=True)
         return {"error": {"message": "Internal server error"}}