Update app.py

app.py

@@ -18,10 +18,11 @@ from huggingface_hub import InferenceClient, login
 
 class BasicAgent:
     def __init__(self):
+        # Required for gated models
         login(token=os.environ["HF_TOKEN"])
         
         self.client = InferenceClient(
-            model="Qwen/Qwen2-7B-Instruct",
+            model="Qwen/Qwen2-7B-Instruct-GPTQ-Int8",
             token=os.environ["HF_TOKEN"],
             timeout=120
         )
@@ -43,12 +44,11 @@ Answer with ONLY the exact value requested.<|im_end|>
                 prompt=prompt,
                 temperature=0.01,
                 max_new_tokens=100,
-                stop_sequences=["<|im_end|>"],
-                repetition_penalty=1.1
+                stop_sequences=["<|im_end|>"]
             )
             return response.split("<|im_start|>assistant")[-1].split("<|im_end|>")[0].strip()
         except Exception as e:
-            print(f"Error processing '{question[:50]}...': {str(e)}")
+            print(f"Error: {str(e)}")
             return ""
 
 def run_and_submit_all( profile: gr.OAuthProfile | None):