From b7d4f85ff8a1dcc79ab75dde186f71acf376ed0b Mon Sep 17 00:00:00 2001
From: "H.Chen" <2366759958@qq.com>
Date: Mon, 26 Mar 2018 16:43:26 +0800
Subject: [PATCH] Update README.md

---
 3-atari/2-pong/README.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/3-atari/2-pong/README.md b/3-atari/2-pong/README.md
index ed36684b..0bb45f69 100644
--- a/3-atari/2-pong/README.md
+++ b/3-atari/2-pong/README.md
@@ -4,10 +4,10 @@ Minimal implementation of Stochastic Policy Gradient Algorithm in Keras
 
 ## Pong Agent
 
-![pg](assetsg.gif)
+![pg](assets/pg.gif)
 
 
 This PG agent seems to get more frequent wins after about 8000 episodes. Below is the score graph.
 
 
-![score](assetscore.png)
+![score](assets/score.png)