use a better model

app.py

@@ -63,12 +63,12 @@ class BasicAgent:
 
         self.llm = ChatHuggingFace(
             llm=HuggingFaceEndpoint(
-                repo_id="meta-llama/Llama-3.1-8B-Instruct",
+                repo_id="deepseek-ai/DeepSeek-R1",
                 task="text-generation",
                 max_new_tokens=512,
                 temperature=0.0,
                 do_sample=False,
-                provider="hf-inference"  # Automatically select the first available provider
+                provider="auto"  # Automatically select the first available provider
             )
         )