diff --git a/.scripts/update_readme_paddle_version.py b/.scripts/update_readme_paddle_version.py
new file mode 100644
index 0000000000000000000000000000000000000000..56d56914c65956a2bb753bc58269d59034766b1c
--- /dev/null
+++ b/.scripts/update_readme_paddle_version.py
@@ -0,0 +1,53 @@
+#   Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import os
+import re
+
+
+def update(fname, ver):
+    with open(fname, 'r') as f:
+        lines = f.readlines()
+
+    for i, line in enumerate(lines):
+        if 'paddlepaddle>=' in line:
+            lines[i] = re.sub("paddlepaddle>=[\d+\.]+",
+                              "paddlepaddle>={}".format(ver), line)
+
+    with open(fname, 'w') as f:
+        for line in lines:
+            f.write(line)
+
+
+if __name__ == '__main__':
+    new_version = '1.6.1'
+
+    readme_files = ['../README.md', '../README.cn.md']
+
+    exclude_examples = [
+        'NeurIPS2019-Learn-to-Move-Challenge',
+        'NeurIPS2018-AI-for-Prosthetics-Challenge', 'EagerMode'
+    ]
+    for example in os.listdir('../examples/'):
+        if example not in exclude_examples:
+            readme_files.append(
+                os.path.join('../examples', example, 'README.md'))
+
+    for example in os.listdir('../examples/EagerMode/'):
+        readme_files.append(
+            os.path.join('../examples/EagerMode', example, 'README.md'))
+
+    print(readme_files)
+    for fname in readme_files:
+        update(fname, new_version)
diff --git a/README.cn.md b/README.cn.md
index dd154a4e634374d5765727e86607e5c178719056..8fc30c5f22bbd0be8751a95f8cf258f5c08fd6de 100644
--- a/README.cn.md
+++ b/README.cn.md
@@ -62,7 +62,7 @@ ans = agent.sum(1,5) # run remotely and not comsume any local computation resour
 # 安装:
 ### 依赖
 - Python 2.7 or 3.5+. 
-- [PaddlePaddle](https://github.com/PaddlePaddle/Paddle) >=1.2.1 (**非必须的**，如果你只用并行部分的接口不需要安装paddle) 
+- [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle) (**非必须的**，如果你只用并行部分的接口不需要安装paddle) 
 
 
 ```
diff --git a/README.md b/README.md
index 7bec38c8a253b4cc53fb4cc0ac0dee3221181b05..93590cfd8af98f10552ade36e9c66c37bc626bc3 100644
--- a/README.md
+++ b/README.md
@@ -65,7 +65,7 @@ For users, they can write code in a simple way, just like writing multi-thread c
 # Install:
 ### Dependencies
 - Python 2.7 or 3.5+. 
-- [PaddlePaddle](https://github.com/PaddlePaddle/Paddle) >=1.5.1 (**Optional**, if you only want to use APIs related to parallelization alone)  
+- [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle) (**Optional**, if you only want to use APIs related to parallelization alone)  
 
 
 ```
diff --git a/examples/A2C/README.md b/examples/A2C/README.md
index bb69f56c52714f0891977d48fb4f2ea84892ad69..d38a5d153b3ab39c59b851775567d90edcdda4fb 100755
--- a/examples/A2C/README.md
+++ b/examples/A2C/README.md
@@ -19,7 +19,7 @@ Performance of A2C on various envrionments
 
 ## How to use
 ### Dependencies
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym==0.12.1
 + atari-py==0.1.7
diff --git a/examples/DDPG/README.md b/examples/DDPG/README.md
index d6274c75403849fef9f4c1828ac200646cb305b6..d7794617a2b93cc98f4ca760d6960ea10defb0cc 100644
--- a/examples/DDPG/README.md
+++ b/examples/DDPG/README.md
@@ -15,7 +15,7 @@ Please see [here](https://github.com/openai/mujoco-py) to know more about Mujoco
 ## How to use
 ### Dependencies:
 + python3.5+
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym
 + tqdm
diff --git a/examples/DQN/README.md b/examples/DQN/README.md
index 4e2fe5d2a893b50a7110ca1a56b1b6b1c7917109..351e44754ad82125eec4e1346fd6301e8c1555b7 100644
--- a/examples/DQN/README.md
+++ b/examples/DQN/README.md
@@ -21,7 +21,7 @@ Performance of DQN on various environments
 
 ## How to use
 ### Dependencies:
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym
 + tqdm
diff --git a/examples/ES/README.md b/examples/ES/README.md
index e1d44d24d949eca8d08b7adfd4e8cc81337d2dbc..207ae2dafa68c5f7d2eb30f956355b07c1bd5d61 100644
--- a/examples/ES/README.md
+++ b/examples/ES/README.md
@@ -12,7 +12,7 @@ Please see [here](https://github.com/openai/mujoco-py) to know more about Mujoco
 
 ## How to use
 ### Dependencies
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym==0.9.4
 + mujoco-py==0.5.1
diff --git a/examples/EagerMode/QuickStart/README.md b/examples/EagerMode/QuickStart/README.md
index b4493ca3920fbf0db53d3fab8104c627e8fbc2cd..fa7612fe5b07b5d07962fa53ff04d39a18bee945 100644
--- a/examples/EagerMode/QuickStart/README.md
+++ b/examples/EagerMode/QuickStart/README.md
@@ -4,7 +4,7 @@ Train an agent with PARL to solve the CartPole problem, a classical benchmark in
 ## How to use
 ### Dependencies:
 
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym
 
diff --git a/examples/GA3C/README.md b/examples/GA3C/README.md
index 5b892fdb6c6108f1dd12b923d5057ddfeaadae3e..220e75fb5340020d5dcc63ec46b9ad3676453711 100755
--- a/examples/GA3C/README.md
+++ b/examples/GA3C/README.md
@@ -16,7 +16,7 @@ Results with one learner (in a P40 GPU) and 24 simulators (in 12 CPU) in 10 mill
 
 ## How to use
 ### Dependencies
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym==0.12.1
 + atari-py==0.1.7
diff --git a/examples/IMPALA/README.md b/examples/IMPALA/README.md
index cd361daf894ea0cfaa5e056204cc938076ff4541..6d96331aa049fc21eae4f288c47eaba6ab48cec6 100755
--- a/examples/IMPALA/README.md
+++ b/examples/IMPALA/README.md
@@ -20,7 +20,7 @@ Result with one learner (in a P40 GPU) and 32 actors (in 32 CPUs).
 
 ## How to use
 ### Dependencies
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym==0.12.1
 + atari-py==0.1.7
diff --git a/examples/MADDPG/README.md b/examples/MADDPG/README.md
index d37b553de416fe9176119f654508774d8bb1278c..55d191474b62f5099d91da51bd80443abd6b87d8 100644
--- a/examples/MADDPG/README.md
+++ b/examples/MADDPG/README.md
@@ -95,7 +95,7 @@ simple_world_comm<br>
 ## How to use
 ### Dependencies:
 + python3.5+
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + [multiagent-particle-envs](https://github.com/openai/multiagent-particle-envs)
 + gym
diff --git a/examples/NeurIPS2018-AI-for-Prosthetics-Challenge/README.md b/examples/NeurIPS2018-AI-for-Prosthetics-Challenge/README.md
index dc5c8aa54fcd1c4a5caf7af1b3a27c20efd92b20..d889335168aa902ae7ba86317ea85c48ede29e32 100644
--- a/examples/NeurIPS2018-AI-for-Prosthetics-Challenge/README.md
+++ b/examples/NeurIPS2018-AI-for-Prosthetics-Challenge/README.md
@@ -19,6 +19,7 @@ For more technical details about our solution, we provide:
 
 ## Dependencies
 - python3.6
+- [parl==1.0](https://github.com/PaddlePaddle/PARL)
 - [paddlepaddle==1.5.1](https://github.com/PaddlePaddle/Paddle)
 - [osim-rl](https://github.com/stanfordnmbl/osim-rl)
 - [grpcio==1.12.1](https://grpc.io/docs/quickstart/python.html)
diff --git a/examples/NeurIPS2019-Learn-to-Move-Challenge/README.md b/examples/NeurIPS2019-Learn-to-Move-Challenge/README.md
index e407c2f0c8f49d7e11ea5a74c1c4ec31eabcb30e..d67fd52f97fef04769501e24a0a802384fe124db 100644
--- a/examples/NeurIPS2019-Learn-to-Move-Challenge/README.md
+++ b/examples/NeurIPS2019-Learn-to-Move-Challenge/README.md
@@ -6,6 +6,7 @@ The **PARL** team gets the first place in NeurIPS reinforcement learning competi
 
 ## Dependencies
 - python3.6
+- [parl==1.2.1](https://github.com/PaddlePaddle/PARL)
 - [paddlepaddle==1.5.1](https://github.com/PaddlePaddle/Paddle)
 - [parl>=1.2.1](https://github.com/PaddlePaddle/PARL)
 - [osim-rl==3.0.11](https://github.com/stanfordnmbl/osim-rl)
diff --git a/examples/PPO/README.md b/examples/PPO/README.md
index cd1bc133031a77242760c57a0a3f8a779be422d8..9a87c1192bbee1a1d30ec434c9f3ddb6b79f726c 100644
--- a/examples/PPO/README.md
+++ b/examples/PPO/README.md
@@ -18,7 +18,7 @@ Please see [here](https://github.com/openai/mujoco-py) to know more about Mujoco
 ## How to use
 ### Dependencies:
 + python3.5+
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym
 + tqdm
diff --git a/examples/QuickStart/README.md b/examples/QuickStart/README.md
index 05cbba67a711665ca08f512674ebd981c353199b..cf11088f7a8d8053f5ae8bc6b5833a148b3c5c89 100644
--- a/examples/QuickStart/README.md
+++ b/examples/QuickStart/README.md
@@ -4,7 +4,7 @@ Train an agent with PARL to solve the CartPole problem, a classical benchmark in
 ## How to use
 ### Dependencies:
 
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym
 
diff --git a/examples/SAC/README.md b/examples/SAC/README.md
index c9a02209d663556900842317aa5f1ab987e14af3..05854e54a339cf44e1157051a575907bf714062c 100644
--- a/examples/SAC/README.md
+++ b/examples/SAC/README.md
@@ -18,7 +18,7 @@ Please see [here](https://github.com/openai/mujoco-py) to know more about Mujoco
 ## How to use
 ### Dependencies:
 + python3.5+
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym
 + mujoco-py>=1.50.1.0
diff --git a/examples/TD3/README.md b/examples/TD3/README.md
index 941ce570d115c5a95ed11c105b27d634440ce5a7..a1aee0cefa4cd1b42dcaf742295ba6f423710b1f 100644
--- a/examples/TD3/README.md
+++ b/examples/TD3/README.md
@@ -19,7 +19,7 @@ Please see [here](https://github.com/openai/mujoco-py) to know more about Mujoco
 ## How to use
 ### Dependencies:
 + python3.5+
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym
 + mujoco-py>=1.50.1.0
diff --git a/examples/offline-Q-learning/README.md b/examples/offline-Q-learning/README.md
index f442f5d9422c753425d05144accf4b0602e78daf..ea63f1fed50c871ca9e4e1d1a7d092e382f32a74 100644
--- a/examples/offline-Q-learning/README.md
+++ b/examples/offline-Q-learning/README.md
@@ -21,7 +21,7 @@ learn_program = parl.compile(learn_program, loss=training_loss)
 We provide a demonstration of offline Q-learning with parallel executing, in which we seperate the procedures of collecting data and training the model. First we collect data by interacting with the environment and save them to a replay memory file, and then fit and evaluate the Q network with the collected data. Repeat these two steps to improve the performance gradually.
 
 ### Dependencies:
-+ [paddlepaddle>=1.5.1](https://github.com/PaddlePaddle/Paddle)
++ [paddlepaddle>=1.6.1](https://github.com/PaddlePaddle/Paddle)
 + [parl](https://github.com/PaddlePaddle/PARL)
 + gym
 + tqdm