diff --git a/python/paddle/trainer_config_helpers/attrs.py b/python/paddle/trainer_config_helpers/attrs.py
index bf0208834600fef3bcf1b0496da8f5f77aea44c5..41ad05c3eb467b9a2e09315f980410d5e9b3853f 100644
--- a/python/paddle/trainer_config_helpers/attrs.py
+++ b/python/paddle/trainer_config_helpers/attrs.py
@@ -196,7 +196,7 @@ class ExtraLayerAttribute(object):
                       <https://www.cs.toronto.edu/~hinton/absps/
                       JMLRdropout.pdf>`_.
     :type drop_rate: float
-    :param device: device ID of layer. device=-1, use CPU. device>0, use GPU.
+    :param device: device ID of layer. device=-1, use CPU. device>=0, use GPU.
                    The details allocation in parallel_nn please refer to `here
                    <http://www.paddlepaddle.org/doc/ui/cmd_argument/
                    use_case.html#case-2-specify-layers-in-different-devices>`_.