diff --git a/demo/mnist/api_train_v2.py b/demo/mnist/api_train_v2.py
index 81b330a60522eaccfa24af5bbed71101aea26f48..d0aca5eb1d94a36a66a2e1c146adb54851baef71 100644
--- a/demo/mnist/api_train_v2.py
+++ b/demo/mnist/api_train_v2.py
@@ -113,7 +113,7 @@ def main():
     test_creator = paddle.dataset.mnist.test()
     test_data = []
     for item in test_creator():
-        test_data.append(item[0])
+        test_data.append((item[0], ))
         if len(test_data) == 100:
             break
 
diff --git a/python/paddle/v2/inference.py b/python/paddle/v2/inference.py
index 7c079a0d32df6f6ae5399b1bba86dcc5db9a19dc..7d7dc82de987cb23d12c411c08e0e529afefe58b 100644
--- a/python/paddle/v2/inference.py
+++ b/python/paddle/v2/inference.py
@@ -43,10 +43,7 @@ class Inference(object):
 
             def __reader_impl__():
                 for each_sample in input:
-                    if len(reader_dict) == 1:
-                        yield [each_sample]
-                    else:
-                        yield each_sample
+                    yield each_sample
 
             reader = minibatch.batch(__reader_impl__, batch_size=batch_size)
         else: