RL关键算法的结构实现_Python_下载.zip

<p align="center"> <img src="https://user-images.githubusercontent.com/17582508/52845370-4a930200-314a-11e9-9889-e00007043872.jpg" align="center"> [](https://lgtm.com/projects/g/medipixel/rl_algorithms/context:python) [](https://opensource.org/licenses/MIT) [](https://github.com/psf/black)<!-- ALL-CONTRIBUTORS-BADGE:START - Do not remove or modify this section --> [](#contributors-) <!-- ALL-CONTRIBUTORS-BADGE:END --> </p> ## Contents * [Welcome!](https://github.com/medipixel/rl_algorithms#welcome) * [Contributors](https://github.com/medipixel/rl_algorithms#contributors) * [Algorithms](https://github.com/medipixel/rl_algorithms#algorithms) * [Performance](https://github.com/medipixel/rl_algorithms#performance) * [Getting Started](https://github.com/medipixel/rl_algorithms#getting-started) * [Class Diagram](https://github.com/medipixel/rl_algorithms#class-diagram) * [References](https://github.com/medipixel/rl_algorithms#references) ## Welcome! This repository contains Reinforcement Learning algorithms which are being used for research activities at Medipixel. The source code will be frequently updated. We are warmly welcoming external contributors! :) |<img src="https://user-images.githubusercontent.com/17582508/52840582-18c76e80-313d-11e9-9752-3d6138f39a15.gif" width="260" height="180"/>|<img src="https://media.giphy.com/media/ZxLNajigOcLyeUnOwg/giphy.gif" width="160" height="180"/>|<img src="https://media.giphy.com/media/1mikGEln2lArKMQ6Pt/giphy.gif" width="260" height="180"/>| |:---:|:---:|:---:| |BC agent on LunarLanderContinuous-v2|RainbowIQN agent on PongNoFrameskip-v4|SAC agent on Reacher-v2| ## Contributors Thanks goes to these wonderful people ([emoji key](https://allcontributors.org/docs/en/emoji-key)): <!-- ALL-CONTRIBUTORS-LIST:START - Do not remove or modify this section --> <!-- prettier-ignore-start --> <!-- markdownlint-disable --> <table> <tr> <td align="center"><a href="https://github.com/Curt-Park"><img src="https://avatars3.githubusercontent.com/u/14961526?v=4?s=100" width="100px;" alt=""/><br />_{<b>Jinwoo Park (Curt)</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=Curt-Park" title="Code">💻</a></td> <td align="center"><a href="https://github.com/MrSyee"><img src="https://avatars3.githubusercontent.com/u/17582508?v=4?s=100" width="100px;" alt=""/><br />_{<b>Kyunghwan Kim</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=MrSyee" title="Code">💻</a></td> <td align="center"><a href="https://github.com/darthegg"><img src="https://avatars3.githubusercontent.com/u/16010242?v=4?s=100" width="100px;" alt=""/><br />_{<b>darthegg</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=darthegg" title="Code">💻</a></td> <td align="center"><a href="https://github.com/mclearning2"><img src="https://avatars3.githubusercontent.com/u/43226417?v=4?s=100" width="100px;" alt=""/><br />_{<b>Mincheol Kim</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=mclearning2" title="Code">💻</a></td> <td align="center"><a href="https://github.com/minseop4898"><img src="https://avatars1.githubusercontent.com/u/34338299?v=4?s=100" width="100px;" alt=""/><br />_{<b>김민섭</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=minseop4898" title="Code">💻</a></td> <td align="center"><a href="https://github.com/jinPrelude"><img src="https://avatars1.githubusercontent.com/u/16518993?v=4?s=100" width="100px;" alt=""/><br />_{<b>Leejin Jung</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=jinPrelude" title="Code">💻</a></td> <td align="center"><a href="https://github.com/cyoon1729"><img src="https://avatars2.githubusercontent.com/u/33583101?v=4?s=100" width="100px;" alt=""/><br />_{<b>Chris Yoon</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=cyoon1729" title="Code">💻</a></td> </tr> <tr> <td align="center"><a href="https://jiseonghan.github.io/"><img src="https://avatars2.githubusercontent.com/u/48741026?v=4?s=100" width="100px;" alt=""/><br />_{<b>Jiseong Han</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=jiseongHAN" title="Code">💻</a></td> <td align="center"><a href="https://github.com/sehyun-hwang"><img src="https://avatars3.githubusercontent.com/u/23437715?v=4?s=100" width="100px;" alt=""/><br />_{<b>Sehyun Hwang</b>}</a><br /><a href="#maintenance-sehyun-hwang" title="Maintenance">🚧</a></td> <td align="center"><a href="https://github.com/isk03276"><img src="https://avatars.githubusercontent.com/u/23740495?v=4?s=100" width="100px;" alt=""/><br />_{<b>eunjin</b>}</a><br /><a href="https://github.com/medipixel/rl_algorithms/commits?author=isk03276" title="Code">💻</a></td> </tr> </table> <!-- markdownlint-restore --> <!-- prettier-ignore-end --> <!-- ALL-CONTRIBUTORS-LIST:END --> This project follows the [all-contributors](https://github.com/all-contributors/all-contributors) specification. ## Algorithms 0. [Advantage Actor-Critic (A2C)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/a2c) 1. [Deep Deterministic Policy Gradient (DDPG)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/ddpg) 2. [Proximal Policy Optimization Algorithms (PPO)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/ppo) 3. [Twin Delayed Deep Deterministic Policy Gradient Algorithm (TD3)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/td3) 4. [Soft Actor Critic Algorithm (SAC)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/sac) 5. [Behaviour Cloning (BC with DDPG, SAC)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/bc) 6. [From Demonstrations (DDPGfD, SACfD, DQfD)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/fd) 7. [Rainbow DQN](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/dqn) 8. [Rainbow IQN (without DuelingNet)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/dqn) - DuelingNet [degrades performance](https://github.com/medipixel/rl_algorithms/pull/137) 9. Rainbow IQN (with [ResNet](https://github.com/medipixel/rl_algorithms/blob/master/rl_algorithms/common/networks/backbones/resnet.py)) 10. [Recurrent Replay DQN (R2D1)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/recurrent) 11. [Distributed Pioritized Experience Replay (Ape-X)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/common/apex) 12. [Policy Distillation](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/distillation) 13. [Generative Adversarial Imitation Learning (GAIL)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/gail) 14. [Sample Efficient Actor-Critic with Experience Replay (ACER)](https://github.com/medipixel/rl_algorithms/tree/master/rl_algorithms/acer) ## Performance We have tested each algorithm on some of the following environments. - [PongNoFrameskip-v4](https://github.com/medipixel/rl_algorithms/tree/master/configs/pong_no_frameskip_v4) - [LunarLanderContinuous-v2](https://github.com/medipixel/rl_algorithms/tree/master/configs/lunarlander_continuous_v2) - [LunarLander_v2](https://github.com/medipixel/rl_algorithms/tree/master/configs/lunarlander_v2) - [Reacher-v2](https://github.com/medipixel/rl_algorithms/tree/master/configs/reacher-v2) ❗Please note that this won't be fr