Upload app.py with huggingface_hub

app.py

@@ -225,8 +225,8 @@ def load_model(model_key: str):
     _current_model = Llama.from_pretrained(
         repo_id=config['repo_id'],
         filename=config['filename'],
-        n_ctx=512,
-        n_threads=4,
+        n_ctx=256,
+        n_threads=8,
         verbose=False,
     )
     _current_model_name = model_key
@@ -276,7 +276,7 @@ Answer:"""
 
     response = model.create_chat_completion(
         messages=messages,
-        max_tokens=20,
+        max_tokens=8,
         temperature=0,
     )