Switch to Groq llama-3.1-8b-instant via LiteLLM

app.py

@@ -67,10 +67,10 @@ CRITICAL RULES:
 """
 
 def build_agent():
-    model = InferenceClientModel(
-        model_id="llama-3.1-8b-instant",
-        token=os.getenv("HF"),
-        max_tokens=2048,
+    model = LiteLLMModel(
+        model_id="groq/llama-3.1-8b-instant",
+        api_key=os.getenv("HF"),
+        max_tokens=1024,
         temperature=0.1,
     )
     agent = CodeAgent(

requirements.txt

@@ -3,3 +3,4 @@ gradio
 requests
 pandas
 duckduckgo-search
+litellm