面向科研和产品化的深度学习和强化学习库.zip

# Comprehensive Reinforcement Learning Tutorial  [](https://github.com/tensorflow/tensorflow/releases) [](https://tensorlayer.readthedocs.io/) [](https://travis-ci.org/tensorlayer/tensorlayer) [](http://pepy.tech/project/tensorlayer) <br/> <a href="https://deepreinforcementlearningbook.org" target="\_blank"> <div align="center"> <img src="http://deep-reinforcement-learning-book.github.io/assets/images/cover_v1.png" width="22%"/> </div> <!-- <div align="center"><caption>Slack Invitation Link</caption></div> --> </a> <br/> <!-- <br/> <a href="https://github.com/tensorlayer/tensorlayer-chinese/blob/master/docs/images/RL_Group_QR.jpeg" target="\_blank"> <div align="center"> <img src="https://github.com/tensorlayer/tensorlayer-chinese/blob/master/docs/images/RL_Group_QR.jpeg" width="20%"/> </div> <div align="center"><caption>WeChat QR Code</caption></div> </a> <br/> --> This repository contains implementations of the most popular reinforcement learning algorithms, powered by [Tensorflow 2.0](https://www.tensorflow.org/alpha/guide/effective_tf2) and Tensorlayer 2.0. We aim to make the reinforcement learning tutorial simple, transparent and straight-forward, as this would not only benefits new learners of reinforcement learning, but also provide convenience for senior researchers to testify their new ideas quickly. A corresponding [Springer textbook](https://deepreinforcementlearningbook.org) is also provided, you can get the free PDF if your institute has Springer license. We also released an [RLzoo](https://github.com/tensorlayer/RLzoo) for simple usage. <br/> <a href="https://join.slack.com/t/tensorlayer/shared_invite/enQtMjUyMjczMzU2Njg4LWI0MWU0MDFkOWY2YjQ4YjVhMzI5M2VlZmE4YTNhNGY1NjZhMzUwMmQ2MTc0YWRjMjQzMjdjMTg2MWQ2ZWJhYzc" target="\_blank"> <div align="center"> <img src="../../img/join_slack.png" width="20%"/> </div> <!-- <div align="center"><caption>Slack Invitation Link</caption></div> --> </a> <br/> ## Prerequisites: * python 3.5 * tensorflow >= 2.0.0 or tensorflow-gpu >= 2.0.0a0 * tensorlayer >= 2.0.1 * tensorflow-probability *** If you meet the error`AttributeError: module 'tensorflow' has no attribute 'contrib'` when running the code after installing tensorflow-probability, try: `pip install --upgrade tf-nightly-2.0-preview tfp-nightly` ## Quick Start ``` conda create --name tl python=3.6.4 conda activate tl pip install tensorflow-gpu==2.0.0-rc1 # if no GPU, use pip install tensorflow==2.0.0 pip install tensorlayer pip install tensorflow-probability==0.9.0 pip install gym pip install gym[atari] # for others, use pip instal gym[all] python tutorial_DDPG.py --train ``` ## Status: Beta We are currently open to any suggestions or pull requests from you to make the reinforcement learning tutorial with TensorLayer2.0 a better code repository for both new learners and senior researchers. Some of the algorithms mentioned in the this markdown may be not yet available, since we are still trying to implement more RL algorithms and optimize their performances. However, those algorithms listed above will come out in a few weeks, and the repository will keep updating more advanced RL algorithms in the future. ## To Use: For each tutorial, open a terminal and run: `python ***.py --train` for training and `python ***.py --test` for testing. The tutorial algorithms follow the same basic structure, as shown in file: [`./tutorial_format.py`](https://github.com/tensorlayer/tensorlayer/blob/reinforcement-learning/examples/reinforcement_learning/tutorial_format.py) The pretrained models and learning curves for each algorithm are stored [here](https://github.com/tensorlayer/pretrained-models). You can download the models and load the weights in the policies for tests. ## Table of Contents: ### value-based | Algorithms | Action Space | Tutorial Env | Papers | | --------------- | ------------ | -------------- | -------| |**value-based**|||| | Q-learning | Discrete | FrozenLake | [Technical note: Q-learning. Watkins et al. 1992](http://www.gatsby.ucl.ac.uk/~dayan/papers/cjch.pdf)| | Deep Q-Network (DQN)| Discrete | FrozenLake | [Human-level control through deep reinforcement learning, Mnih et al. 2015.](https://www.nature.com/articles/nature14236/) | | Prioritized Experience Replay | Discrete | Pong, CartPole | [Schaul et al. Prioritized experience replay. Schaul et al. 2015.](https://arxiv.org/abs/1511.05952) | |Dueling DQN|Discrete | Pong, CartPole |[Dueling network architectures for deep reinforcement learning. Wang et al. 2015.](https://arxiv.org/abs/1511.06581)| |Double DQN| Discrete | Pong, CartPole |[Deep reinforcement learning with double q-learning. Van et al. 2016.](https://arxiv.org/abs/1509.06461)| |Noisy DQN|Discrete | Pong, CartPole |[Noisy networks for exploration. Fortunato et al. 2017.](https://arxiv.org/pdf/1706.10295.pdf)| | Distributed DQN (C51)| Discrete | Pong, CartPole | [A distributional perspective on reinforcement learning. Bellemare et al. 2017.](https://arxiv.org/pdf/1707.06887.pdf) | |**policy-based**|||| |REINFORCE(PG) |Discrete/Continuous|CartPole | [Reinforcement learning: An introduction. Sutton et al. 2011.](https://www.cambridge.org/core/journals/robotica/article/robot-learning-edited-by-jonathan-h-connell-and-sridhar-mahadevan-kluwer-boston-19931997-xii240-pp-isbn-0792393651-hardback-21800-guilders-12000-8995/737FD21CA908246DF17779E9C20B6DF6)| | Trust Region Policy Optimization (TRPO)| Discrete/Continuous | Pendulum | [Abbeel et al. Trust region policy optimization. Schulman et al.2015.](https://arxiv.org/pdf/1502.05477.pdf) | | Proximal Policy Optimization (PPO) |Discrete/Continuous |Pendulum| [Proximal policy optimization algorithms. Schulman et al. 2017.](https://arxiv.org/abs/1707.06347) | |Distributed Proximal Policy Optimization (DPPO)|Discrete/Continuous |Pendulum|[Emergence of locomotion behaviours in rich environments. Heess et al. 2017.](https://arxiv.org/abs/1707.02286)| |**actor-critic**|||| |Actor-Critic (AC)|Discrete/Continuous|CartPole| [Actor-critic algorithms. Konda er al. 2000.](https://papers.nips.cc/paper/1786-actor-critic-algorithms.pdf)| | Asynchronous Advantage Actor-Critic (A3C)| Discrete/Continuous | BipedalWalker| [Asynchronous methods for deep reinforcement learning. Mnih et al. 2016.](https://arxiv.org/pdf/1602.01783.pdf) | | DDPG|Discrete/Continuous |Pendulum| [Continuous Control With Deep Reinforcement Learning, Lillicrap et al. 2016](https://arxiv.org/pdf/1509.02971.pdf) | |TD3|Discrete/Continuous |Pendulum|[Addressing function approximation error in actor-critic methods. Fujimoto et al. 2018.](https://arxiv.org/pdf/1802.09477.pdf)| |Soft Actor-Critic (SAC)|Discrete/Continuous |Pendulum|[Soft actor-critic algorithms and applications. Haarnoja et al. 2018.](https://arxiv.org/abs/1812.05905)| ## Examples of RL Algorithms: * **Q-learning** Code: `./tutorial_Qlearning.py` <u>Paper</u>: [Technical Note Q-Learning](http://www.gatsby.ucl.ac.uk/~dayan/papers/cjch.pdf) <u>Description</u>: ``` Q-learning is a non-deep-learning method with TD Learning, Off-Policy, e-Greedy Exploration. Central formula: Q(S, A) <- Q(S, A) + alpha * (R + lambda * Q(newS, newA) - Q(S, A)) See David Silver RL Tutorial Lecture 5 - Q-Learning for more details. ``` * **Deep Q-Network (DQN)** <u>Code:</u> `./tutorial_DQN.py` <u>Paper</u>: [Human-level control through deep reinforcementlearning](https://