fix

ai_service/api.py

@@ -7,6 +7,8 @@ from fastapi import FastAPI, HTTPException
 from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel, Field
 from typing import List, Optional
+import uuid
+from datetime import datetime, timezone
 
 from ai_service.models.model_service import model_service
 from ai_service.models.chat_service import chat_service
@@ -120,7 +122,7 @@ async def test_ollama_connection():
 
     try:
         # Try to connect to Ollama API
-        response = requests.get(f"{config.OLLAMA_API_URL}/api/tags", timeout=5)
+        response = requests.get(f"{config.OLLAMA_API_URL}/api/tags", timeout=config.API_TIMEOUT)
         response.raise_for_status()
 
         # Return the models from Ollama
@@ -130,6 +132,106 @@ async def test_ollama_connection():
             "ollama_url": config.OLLAMA_API_URL,
             "models": response.json()
         }
+    except requests.exceptions.Timeout as e:
+        return {
+            "status": "error",
+            "message": f"Timeout connecting to Ollama API: {str(e)}. The request exceeded the {config.API_TIMEOUT} second timeout.",
+            "ollama_url": config.OLLAMA_API_URL
+        }
+    except requests.exceptions.ConnectionError as e:
+        return {
+            "status": "error",
+            "message": f"Connection error to Ollama API: {str(e)}. Please check if Ollama is running at {config.OLLAMA_API_URL}.",
+            "ollama_url": config.OLLAMA_API_URL
+        }
+    except Exception as e:
+        return {
+            "status": "error",
+            "message": f"Failed to connect to Ollama API: {str(e)}",
+            "ollama_url": config.OLLAMA_API_URL
+        }
+
+@app.post("/test-chat")
+async def test_chat_completion():
+    """
+    Test the chat completion with a simple prompt.
+
+    Returns:
+        Model response.
+    """
+    try:
+        # Use the model service directly
+        response = model_service.generate_response(
+            model_id=config.DEFAULT_MODEL,
+            prompt="Hello, how are you?",
+            context=[],
+            use_rag=False
+        )
+
+        return {
+            "status": "success",
+            "model": config.DEFAULT_MODEL,
+            "response": response,
+            "ollama_url": config.OLLAMA_API_URL
+        }
+    except Exception as e:
+        return {
+            "status": "error",
+            "message": f"Failed to get chat completion: {str(e)}",
+            "ollama_url": config.OLLAMA_API_URL
+        }
+
+@app.post("/test-ollama-direct")
+async def test_ollama_direct():
+    """
+    Test the Ollama API directly with a simple chat request.
+
+    Returns:
+        Raw Ollama API response.
+    """
+    import requests
+
+    try:
+        # Prepare a simple chat request
+        request_json = {
+            "model": config.DEFAULT_MODEL,
+            "messages": [
+                {"role": "system", "content": "You are a helpful assistant."},
+                {"role": "user", "content": "Hello, how are you?"}
+            ],
+            "stream": False
+        }
+
+        # Make the API call to Ollama
+        print(f"Sending direct request to Ollama API at: {config.OLLAMA_API_URL}/api/chat")
+        response = requests.post(
+            f"{config.OLLAMA_API_URL}/api/chat",
+            headers={"Content-Type": "application/json"},
+            json=request_json,
+            timeout=config.API_TIMEOUT
+        )
+
+        response.raise_for_status()
+        result = response.json()
+
+        return {
+            "status": "success",
+            "ollama_url": config.OLLAMA_API_URL,
+            "request": request_json,
+            "response": result
+        }
+    except requests.exceptions.Timeout as e:
+        return {
+            "status": "error",
+            "message": f"Timeout connecting to Ollama API: {str(e)}. The request exceeded the {config.API_TIMEOUT} second timeout.",
+            "ollama_url": config.OLLAMA_API_URL
+        }
+    except requests.exceptions.ConnectionError as e:
+        return {
+            "status": "error",
+            "message": f"Connection error to Ollama API: {str(e)}. Please check if Ollama is running at {config.OLLAMA_API_URL}.",
+            "ollama_url": config.OLLAMA_API_URL
+        }
     except Exception as e:
         return {
             "status": "error",
@@ -151,6 +253,7 @@ async def get_config():
         "openwebui_url": config.OPENWEBUI_URL,
         "ollama_api_url": config.OLLAMA_API_URL,
         "default_model": config.DEFAULT_MODEL,
+        "api_timeout": config.API_TIMEOUT,
         "available_models": list(model_service.AVAILABLE_MODELS.keys())
     }
 
@@ -251,6 +354,11 @@ async def send_message(chat_id: str, request: MessageRequest):
         Bot response message.
     """
     try:
+        print(f"Processing message for chat {chat_id} from user {request.user_id}")
+        print(f"Message: {request.message[:50]}...")  # Print first 50 chars of message
+        print(f"Using RAG: {request.use_rag}")
+        print(f"Model parameters: temperature={request.temperature}, max_tokens={request.max_tokens}")
+
         # Extract model parameters from the request
         response = chat_service.get_chat_response(
             chat_id=chat_id,
@@ -270,10 +378,25 @@ async def send_message(chat_id: str, request: MessageRequest):
             function_calling=request.function_calling
         )
 
+        print(f"Response received. Length: {len(response.get('content', ''))}")
         return response
 
     except ValueError as e:
-        raise HTTPException(status_code=404, detail=str(e))
+        error_msg = f"Chat not found: {str(e)}"
+        print(f"ERROR: {error_msg}")
+        raise HTTPException(status_code=404, detail=error_msg)
+    except Exception as e:
+        error_msg = f"Error processing message: {str(e)}"
+        print(f"ERROR: {error_msg}")
+        # Return an error message instead of raising an exception
+        # This ensures the client gets a proper response
+        return {
+            "id": str(uuid.uuid4()),
+            "content": f"Error processing message: {str(e)}",
+            "user_id": None,
+            "is_user_message": False,
+            "timestamp": datetime.now(timezone.utc).isoformat()
+        }
 
 @app.post("/chats/{chat_id}/members/{user_id}")
 async def add_team_member(chat_id: str, user_id: str):