new llama test

app/main.py

@@ -180,7 +180,7 @@ def get_llm(model_name: str):
         try:
             # Try smaller Llama model that's under the 10GB limit
             return HFInferenceLLM(
-                model="TinyLlama/TinyLlama-1.1B-Chat-v1.0",
+                model="meta-llama/Llama-3.1-8B",
                 provider="hf-inference",
                 api_key=hf_token,
                 max_tokens=4000  # Increased token limit