Please note that D2RL here is not abbreviation for Dense Deep Reinforcement Learning. 请注意这里的D2RL，它不是Dense Deep Reinforcement Learning的缩写。
I am an enthusiast of artificial intelligence and also a successor of the Communism（You should really re-examine and try to achieve the goal pf communism in this era of AI.） 我是一名热衷于人工智能的爱好者，同时也是共产主义事业的接班人。（你真应该在这个AI时代重新了解和试着完成共产主义的目标）
I'm also one of the authors of the book "Dive into Deep Learning".（Past ID:@StevenJokes(not existed)，Now:@StevenJokess：https://github.com/StevenJokess/）我是《动手学深度学习》作者之一（过去ID为@StevenJokes（已不存在），现在为@StevenJokess：https://github.com/StevenJokess/）
Main contributions of the book include: Converting GAN/DCGAN models in last chapter from MXNet to PyTorch/TF2 firstly and a series of learning notes and discussions. 主要贡献是该书最后一章的GAN\DCGAN的MXNet->PyTorch\TF2，DCGAN的MXNet->PyTorch\TF2）和一系列的学习笔记和讨论。
Prove 证明：
- Contribution 贡献：https://github.com/StevenJokess/d2l-en-read/tree/master/My%20contribution
- notes 笔记：https://github.com/StevenJokess/d2l-en-read
- Discussion 讨论：https://discuss.d2l.ai/u/stevenjokes/summary
- Discussion 讨论：https://discuss.d2l.ai/u/stevenjokess/summary
- Resume(with face) 简历（带头像）：https://stevenjokess.github.io/2bPM/get_started.html
This project is a synthesis of my hands-on learning from a series of projects, including "Hands-on Reinforcement Learning," "EasyRL," and other (deep) reinforcement learning resources that I was able to find and understand on the internet. 此项目是我再去动手学 《动手学强化学习》、EasyRL等一系列项目和浏览理解了互联网所有我能找到的（深度）强化学习资料后，综合而成的项目。
Innovation points 创新点有：
- A Tutorial Anyone Can Learn From 傻瓜都能学会的教程！
- Comprehensive historic experimental records 全面的历史试验记录！
- Addition of a large amount of content 增加大量篇幅的内容！
- detailed and smooth introduction to algorithms 详尽、顺畅的算法介绍
- the forefront of RL（DRL, MARL, MADRL,Distributed RL，unsupervised RL，Meta RL）前沿发展，含深度强化学习、多智能体强化学习、多智能体深度强化学习、无监督强化学习、元强化学习...
- mathematical foundations 数学基础
- cutting-edge applications 前沿应用
- my nonsense thinkings 笔者的一些胡说八道式的思考
- detailed reference materials 详尽的参考资料
- Chinese and English content 中英双语
- repo's multiple iteration process visualization 代码库多次迭代过程可视化
- unlike a book need to be pulished, it is forever unfinished and never perfect 不像需要被出版的书，它永不完结、永不完美
- TODO：use Tianshou DRL repo to train 用Tianshou库来训练
This project and its content are welcome to be forked and reposted, but the link must be placed in the first position, otherwise it will be considered as riding on the popularity. If commercial use is intended, please contact (email: [email protected]) for approval before use, otherwise, there will be consequences. This project currently has no other commercial intentions, and only accepts help that can provide me with enough necessary goods and services for living. 本项目及其内容
欢迎fork、转载，但转载时必须将链接放在第一的位置，否则必去蹭热度。若进行商用必先咨询（邮箱：[email protected]）后同意才可使用，否则必去叨扰。本项目暂时无其他商业模式的意向，只接受帮我获得足够生活所需物品和服务的帮助。
This project also aims to explore the extent to which deep reinforcement learning can be achieved solely through open and free resources in this era.
本项目也想尝试在这个时代仅仅凭借公开且免费的资源能把深度强化学习学到什么程度。
Dense Deep Reinforcement Learning
My mission is to ensure that artificial general intelligence (AGI) benefits me a poor guy to have all the necessary things and services for living at least, pursue my passions freely, and have a happy life first, then same to all poors in my mainland China, lastly same to all poors in the world. 我的使命是确保人工智能通用智能（AGI）首先使我，一个穷人，至少拥有所有生活所需物品和服务，自由地追求自己的激情，拥有幸福生活。然后是让祖国所有贫困人口也能如此，最后是让全球所有贫困人口也能如此。
Talking about AGI, virtuously "safety" AGI, I'm not sure that poors like me can handle it and be benefited from it. 说到AGI，伪善的说辞是"安全"AGI，我不确定像我这样的穷人能否应对并从中受益。
It is obvious that now lots of company utilize AI technology only to reduce staff and save labor costs, rather than helping poor people to get enough necessary goods and free to any dream career. 显然，现在很多公司都利用人工智能技术只为了裁员和节省劳动力成本，而不是帮助穷人得到充足的生活所需的物品和服务和拥有选择任何梦想的造业的自由。
I had only about 300 CNY income last three years. 我过去三年大约只有300元人民币的收入。
The donation Bitcoin address / 捐赠比特币地址：I don't have any Bitcoins. Fxxk any Shitcoins（They are only useful to financial speculator and made GPUs expensive）!/我没比特币，去他妈的屎币！（他们只对金融投机家有用，还导致了GPUs昂贵）
You can star or fork this repo as a payment to my Communism work.(The only requirement is that you can't delete this Readme.md file's content!) 你可以star和fork这个代码库作为我共产主义工作的回报。(唯一要求是你不能删除该文件的已有内容！)
You can also email me to have a friendly chat and maybe kindly help me get some “foreign” things or service I need in life. 你也可以邮箱联系我，来友善交流，有可能的话友善地帮我搞点生活所需的“外国”货或服务。
My email: [email protected]
问：本项目是骗钱吗？
- 首先本项目主要是由我自己与家里存款赞助我自己。我只是出于兴趣做的，对深度强化学习强大的兴趣所致的作品。为了要反抗同人的真正活动和本质属性相背离相割裂的“资本”，为了我继续符合我本质的活动，顺便也帮更多人做出符合其自由的活动。钱目前是为了尽可能cover掉创造该项目时的我的生活成本，使我不用在该项目完成前被迫去从事一项短期能赚钱的劳动，目前当然没cover掉，当然能帮我实现项目完成后的自由就更好不过了。
- 再谈资本主义下的“收入”（我对货币的需要，正是由于所谓“占有”我所需要的物品对其的需要。越是持资本主义的观念的人，越会怼那些坚信共产主义的要追求美好生活、甚至仅仅维持生活的物质需要的钱的活动，仿佛必须要我们街头饿死才得逞，要知道我们的惨境正是这些观念所造成的，而我们要争取的也是你们的自由和幸福啊！），马上就要被赶出来了！"正如马克思所说，人们首先必须吃、喝、住、穿，然后才能从事政治、科学、艺术、宗教等等；"（：祝福共产主义快点实现！而由于某些物质的有限性、生产的效率低、人为设定的门槛（专利、秘方），生产资料和所需品常被少数人所控制。
- 于是放上收款码，如果对你有任何帮助、或仅仅为了帮助我，请支持一二，在家"失业"三年，虽然包吃住但是全职被骂，也不好受。
- 如果愿意可以先交个朋友，可＋QQ群 171097552（表面讨论AI、PM和共产主义，实则群主要饭群：[要饭日记](./%E8%A6%81%E9%A5%AD%E6%97%A5%E8%AE%B0.md），也可以邮箱联系我：[email protected]
- 当当代的恩格斯，自愿送点你的闲钱，可以慷慨赞助我但不要影响你自由的那种：）或者以其他方式（很多你认为是废品的东西，对我却是宝贝，欢迎邮寄）赞助我。
- 该项目收入我也会尽量公开，如果生活所需满足了，也就不需要赞助了，我就会撤下收款码。目前收入10元左右。
- 为啥只有产品和服务才能卖，我们的每句话作为信息和数据，不能卖吗？凭什么免费给Github或腾讯公司？
To cite this repo in publications:, please use this bibtex entry:
@misc{d2rl2023,
author = {StevenJokess（蔡舒起，原StevenJokes）},
title = {Dive into Deep Reinforcement Learning},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/StevenJokess/d2rl}},
year = {2023}
}
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent