Update app.py

app.py

@@ -120,7 +120,7 @@ model = TransformerWrapper(
                               )
 )
 
-model = AutoregressiveWrapper(model, ignore_index=PAD_IDX, pad_value=PAD_IDX)
+model = AutoregressiveWrapper(model, ignore_index=PAD_IDX, pad_value=PAD_IDX, return_cache=True)
 
 print('=' * 70)
 print('Loading model checkpoint...')