From 650490a9367d4d04e65fd863e6f448377b1864da Mon Sep 17 00:00:00 2001
From: huangjun12 <2399845970@qq.com>
Date: Fri, 15 Jan 2021 14:19:59 +0800
Subject: [PATCH] Update accelerate.md

---
 docs/zh-CN/tutorials/accelerate.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/zh-CN/tutorials/accelerate.md b/docs/zh-CN/tutorials/accelerate.md
index 683b5079b..c9a178503 100644
--- a/docs/zh-CN/tutorials/accelerate.md
+++ b/docs/zh-CN/tutorials/accelerate.md
@@ -65,7 +65,7 @@ num_workers=4
 基于以上思想，FAIR在实验的基础上提出了Multigrid训练策略: 固定`N*C*T*H*W`的值，降低`T*H*W`时增大`N`的值，增大`T*H*W`时减小`N`的值。具体的有两种策略，如示意图所示：
 
 Long cycle:
-设完整训练需要N个epoch，将整个训练过程分4个阶段，每个阶段分别训练`[N/8, N/4，N/2, N]/(1/8+2/8+4/8+8/8) = [N/15, 2N/15, 4N/15, 8N/15]`个epoch数，每个阶段对应的输入tensor形状为:
+设完整训练需要N个epoch，将整个训练过程分4个阶段，每个阶段对应的输入tensor形状为:
 `[8N, T/4, H/sqrt(2), W/sqrt(2)], [4N, T/2, H/sqrt(2), W/sqrt(2)], [2N, T/2, H, W], [N, T, H, W]`
 
 Short cycle: