Update app.py

app.py

@@ -38,12 +38,13 @@ Answer:"""
         inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
 
         output = model.generate(
-            **inputs,
-            max_new_tokens=256,
-            do_sample=True,
-            top_p=0.9,
-            temperature=0.7,
-            repetition_penalty=1.1
+        **inputs,
+        max_new_tokens=512,   # allow longer answers
+        do_sample=True,
+        top_p=0.9,
+        temperature=0.7,
+        repetition_penalty=1.1
+
         )
 
         decoded_output = tokenizer.decode(output[0], skip_special_tokens=True)