Pass along the raw parameter

The `raw` parameter tells the LLM to never use natural language, but
just reply in the format of the message. We need to pass that to the
generate call or else we migth get garbage back to the client.

Author: jhrozek

src/codegate/providers/ollama/completion_handler.py

@@ -100,6 +100,7 @@ async def execute_completion(
             response = await self.client.generate(
                 model=request["model"],
                 prompt=prompt,
+                raw=request.get("raw", False),
                 suffix=request.get("suffix", ""),
                 stream=stream,
                 options=request["options"],  # type: ignore

tests/providers/ollama/test_ollama_completion_handler.py

@@ -41,6 +41,7 @@ async def test_execute_completion_is_fim_request(handler, chat_request):
         stream=False,
         options=chat_request["options"],
         suffix="",
+        raw=False,
     )