diff --git a/tutorials/source_zh_cn/advanced/modules/loss.ipynb b/tutorials/source_zh_cn/advanced/modules/loss.ipynb
index f7d210beceae21eddadd5ff83c9c8d69bc0d7e4d..307fdd3aa1f6ee9c2fdd1b59c69eaf1846807d15 100644
--- a/tutorials/source_zh_cn/advanced/modules/loss.ipynb
+++ b/tutorials/source_zh_cn/advanced/modules/loss.ipynb
@@ -37,7 +37,7 @@
     "            \\operatorname{sum}(L),  & \\text{if reduction} = \\text{'sum'.}\n",
     "        \\end{cases}$$\n",
     "\n",
-    "`nn.L1Loss`中的参数`reduction`取值可为`mean`，`sum`，或`none`。若`reduction` 为`mean`或`sum`，则输出一个经过均值或求和后的标量Tensor（降维）；若`reduction`为`none`，则所输出Tensor的shape为广播后的shape。"
+    "`nn.L1Loss`中的参数`reduction`取值可为`mean`，`sum`，或`none`，默认值为`mean`。若`reduction` 为`mean`或`sum`，则输出一个经过均值或求和后的标量Tensor（降维）；若`reduction`为`none`，则所输出Tensor的shape为广播后的shape。"
    ]
   },
   {