use the smallest possible model as the default_model

exo/api/chatgpt_api.py

@@ -163,7 +163,7 @@ class ChatGPTAPI:
     self.prompts: PrefixDict[str, PromptSession] = PrefixDict()
     self.prev_token_lens: Dict[str, int] = {}
     self.stream_tasks: Dict[str, asyncio.Task] = {}
-    self.default_model = "llama-3.1-8b"
+    self.default_model = "llama-3.2-1b"
 
     cors = aiohttp_cors.setup(self.app)
     cors_options = aiohttp_cors.ResourceOptions(