Fix text errors in the tutorials (#7959)

* Fix text errors in the tutorials

* Add a blank line.

Author: Chenastron

docs/en/tutorials/customize_losses.md

@@ -6,7 +6,7 @@ This tutorial first elaborate the computation pipeline of losses, then give some
 
 ## Computation pipeline of a loss
 
-Given the input prediction and target, as well as the weights, a loss function maps the input tensor to the final loss scalar. The mapping can be divided into four steps:
+Given the input prediction and target, as well as the weights, a loss function maps the input tensor to the final loss scalar. The mapping can be divided into five steps:
 
 1. Set the sampling method to sample positive and negative samples.
 

docs/zh_cn/tutorials/customize_losses.md

@@ -7,7 +7,8 @@ MMDetection 为用户提供了不同的损失函数。但是默认的配置可
 
 ## 一个损失的计算过程
 
-给定输入（包括预测和目标，以及权重），损失函数会把输入的张量映射到最后的损失标量。映射过程可以分为下面四个步骤：
+给定输入（包括预测和目标，以及权重），损失函数会把输入的张量映射到最后的损失标量。映射过程可以分为下面五个步骤：
+
 1. 设置采样方法为对正负样本进行采样。
 
 2. 通过损失核函数获取**元素**或者**样本**损失。