Update OllamaProvider to use chat endpoint for consistency and fix README config

core/llm.py

@@ -42,10 +42,11 @@ class OllamaProvider(LLMProvider):
 
     def generate(self, prompt: str, max_tokens: int = 500, stream: bool = False) -> Union[str, Generator[str, None, None]]:
         def _make_request():
-            url = f"{self.host}/api/generate"
+            # Use the chat endpoint instead of generate for better compatibility
+            url = f"{self.host}/api/chat"
             payload = {
                 "model": self.model_name,
-                "prompt": prompt,
+                "messages": [{"role": "user", "content": prompt}],
                 "stream": stream,
                 "options": {
                     "num_predict": max_tokens
@@ -69,13 +70,19 @@ class OllamaProvider(LLMProvider):
                         if line:
                             try:
                                 data = json.loads(line.decode('utf-8'))
-                                if 'response' in data:
-                                    yield data['response']
+                                # Handle chat endpoint response format
+                                if 'message' in data and 'content' in data['message']:
+                                    yield data['message']['content']
                             except:
                                 continue
                 return stream_response()
             else:
-                return response.json()["response"]
+                # Handle chat endpoint response format
+                data = response.json()
+                if 'message' in data and 'content' in data['message']:
+                    return data['message']['content']
+                else:
+                    raise Exception("Unexpected response format from Ollama")
         
         # Fixed: Moved return outside the _make_request function
         return self._retry_request(_make_request)