Update app.py

app.py

@@ -59,7 +59,7 @@ def predict(message, history):
     #         break
     #     yield partial_message
     inputs = tokenizer(message, return_tensors="pt")["input_ids"].to(device)
-    outputs = model.generate(inputs, do_sample=False, max_length=100)
+    outputs = model.generate(inputs, do_sample=False, max_length=500)
     print(tokenizer.decode(outputs[0]))
     return(tokenizer.decode(outputs[0]))