Update app.py

app.py

@@ -34,27 +34,18 @@ def llm_chat_response(text: str) -> str:
             raise HTTPException(status_code=500, detail="HF_TOKEN not configured")
         
         logger.info("Initializing InferenceClient...")
-        client = InferenceClient(token=HF_TOKEN)  # Changed from api_key to token
-        
-        messages = [
-            {
-                "role": "user",
-                "content": [
-                    {
-                        "type": "text",
-                        "text": text + str(' describe in one line only')
-                    }
-                ]
-            }
-        ]
+        client = InferenceClient(token=HF_TOKEN)
         
         logger.info("Sending request to model...")
-        response_from_llama = client.chat.completions.create(
+        # Using text-generation endpoint instead of chat
+        response = client.text_generation(
+            text + str(' describe in one line only'),
             model="meta-llama/Llama-3.2-11B-Vision-Instruct",
-            messages=messages,
-            max_tokens=500
+            max_new_tokens=500,
+            temperature=0.7
         )
-        return response_from_llama.choices[0].message['content']
+        return response
+
     except Exception as e:
         logger.error(f"Error in llm_chat_response: {str(e)}")
         raise HTTPException(status_code=500, detail=str(e))