changed the max tokens

max_new_tokens=192 oli 384

app.py

@@ -31,7 +31,7 @@ def generate(question: str, level: str, step_by_step: bool) -> str:
     with torch.no_grad():
         out = model.generate(
             **inputs,
-            max_new_tokens=384,
+            max_new_tokens=192,
             do_sample=True,
             temperature=0.7,
             top_p=0.95,