From 0f998bc2f3c364bdb0944045a89438009fe91d97 Mon Sep 17 00:00:00 2001
From: skindhu <skindhu@tencent.com>
Date: Sun, 27 Oct 2024 11:25:48 +0800
Subject: [PATCH] add third chapter

---
 Book/3.实现注意力机制.md | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/Book/3.实现注意力机制.md b/Book/3.实现注意力机制.md
index 9f8accd..8d44d0e 100644
--- a/Book/3.实现注意力机制.md
+++ b/Book/3.实现注意力机制.md
@@ -284,12 +284,12 @@ Sum: tensor(1.)
 >
 > 1. **Softmax 的原理**
 >
->    Softmax 函数的公式如下：  
+>    Softmax 函数的公式如下：  $\operatorname{softmax}\left(z_{i}\right)=\frac{e^{z_{i}}}{\sum_{j} e^{z_{j}}}$
 >
 >    
->    $$
->    \operatorname{softmax}\left(z_{i}\right)=\frac{e^{z_{i}}}{\sum_{j} e^{z_{j}}}
->    $$
+>
+>    g
+>
 >    This is an inline formula: $E=mc^2$.
 >
 >