From 160aa64132e5d6a371e01bf55eca668a5601c0c7 Mon Sep 17 00:00:00 2001
From: typhoonzero <typhoonzero1986@gmail.com>
Date: Mon, 22 Jan 2018 16:31:20 +0800
Subject: [PATCH] follow comments

---
 doc/faq/local/index_cn.rst | 1 +
 1 file changed, 1 insertion(+)

diff --git a/doc/faq/local/index_cn.rst b/doc/faq/local/index_cn.rst
index d0cdba90d89..0306b1e5dd2 100644
--- a/doc/faq/local/index_cn.rst
+++ b/doc/faq/local/index_cn.rst
@@ -235,6 +235,7 @@ mini-batch forward的output的值。获得的值类型均为 :code:`numpy.ndarra
                                 "average absolute diff scores: %.6f") %
                                 (event.pass_id, event.batch_id, event.cost, diff))
 
+注意：此方法不能获取 :code:`paddle.layer.recurrent_group` 里step的内容，但可以获取 :code:`paddle.layer.recurrent_group` 的输出。
 
 7.  如何在训练过程中获得参数的权重和梯度
 -----------------------------------------------
-- 
GitLab