Update app.py

app.py

@@ -33,7 +33,7 @@ def respond(query):
     eval_prompt = "###Input: " + query + "\n\n###Output: "
     model_input = tokenizer(eval_prompt, return_tensors="pt")#.to("cuda")
     output = ft_model.generate(input_ids=model_input["input_ids"]#.to(device),
-                           attention_mask=model_input["attention_mask"], 
+                           ,attention_mask=model_input["attention_mask"], 
                            max_new_tokens=500)
     result = tokenizer.decode(output[0], skip_special_tokens=True).replace(eval_prompt, "")
     return result