Shunyu Liu / 刘顺宇

Shunyu Liu / 刘顺宇

Ph.D. Student @ ZJU

Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Posts

ssh 密钥登陆服务器

1 minute read

Published:

首先进行 ssh 设置密钥，然后在本地配置 ssh config 文件实现快捷登陆。

强化学习论文（17）MAAC

less than 1 minute read

Published:

标签： MAAC; actor-critic; off-policy; model-free; communication; continuous communication channel; discrete action space; continuous state space; mixed task; cooperative task; competitive task; centralized training with decentralized execution; multi-agent;

强化学习论文（16）Qatten

less than 1 minute read

Published:

标签： Qatten; value-based; off-policy; model-free; discrete action space; continuous state space; cooperative task; credit assignment; centralized training with decentralized execution; value function factorization; multi-agent;

强化学习论文（15）Lenient-DQN

less than 1 minute read

Published:

标签： Lenient-DQN; value-based; off-policy; model-free; discrete action space; continuous state space; cooperative task; decentralized approach; multi-agent;

强化学习论文（14）Dec-HDRQNs&CERTs

1 minute read

Published:

标签： Dec-HDRQNs; CERTs; value-based; off-policy; model-free; discrete action space; continuous state space; cooperative task; decentralized approach; multi-agent; multi-task; transfer learning; distillation;

强化学习论文（13）Fingerprints

less than 1 minute read

Published:

标签： Fingerprints; value-based; off-policy; model-free; discrete action space; continuous state space; cooperative task; decentralized approach; multi-agent;

强化学习论文（12）SchedNet

less than 1 minute read

Published:

标签： SchedNet; actor-critic; off-policy; model-free; communication; continuous communication channel; discrete action space; continuous state space; cooperative task; centralized training with decentralized execution; multi-agent;

强化学习论文（11）ATOC

1 minute read

Published:

标签： ATOC; actor-critic; off-policy; model-free; communication; continuous communication channel; continuous action space; continuous state space; cooperative task; centralized training with decentralized execution; multi-agent;

强化学习论文（10）QTRAN

less than 1 minute read

Published:

标签： QTRAN; value-based; off-policy; model-free; discrete action space; continuous state space; cooperative task; credit assignment; centralized training with decentralized execution; value function factorization; multi-agent;

强化学习论文（9）QMIX

less than 1 minute read

Published:

标签： QMIX; value-based; off-policy; model-free; discrete action space; continuous state space; cooperative task; credit assignment; centralized training with decentralized execution; value function factorization; multi-agent;

强化学习论文（8）VDN

less than 1 minute read

Published:

标签： VDN; value-based; off-policy; model-free; communication; continuous communication channel; discrete action space; continuous state space; cooperative task; credit assignment; centralized training with decentralized execution; value function factorization; multi-agent;

强化学习论文（7）COMA

less than 1 minute read

Published:

标签： COMA; actor-critic; on-policy; model-free; communication; continuous communication channel; discrete action space; continuous state space; cooperative task; credit assignment; centralized training with decentralized execution; multi-agent;

强化学习论文（6）DCC-MD&MADDPG-MD

1 minute read

Published:

标签： DCC-MD; value-based; discrete action space; decentralized approach; MADDPG-MD; actor-critic; continuous action space; centralized training with decentralized execution; off-policy; model-free; communication; continuous communication channel; continuous state space; cooperative task; multi-agent;

强化学习论文（5）MD-MADDPG

less than 1 minute read

Published:

标签： MD-MADDPG; actor-critic; off-policy; model-free; communication; continuous communication channel; continuous action space; continuous state space; cooperative task; centralized training with decentralized execution; multi-agent;

强化学习论文（4）BiCNet

less than 1 minute read

Published:

标签： BiCNet; actor-critic; off-policy; model-free; communication; continuous communication channel; continuous action space; continuous state space; cooperative task; centralized approach; multi-agent;

强化学习论文（3）CommNet

less than 1 minute read

Published:

标签： CommNet; policy-based; on-policy; model-free; communication; continuous communication channel; discrete action space; continuous state space; cooperative task; centralized approach; multi-agent;

强化学习论文（2）RIAL&DIAL

1 minute read

Published:

标签： RIAL; DIAL; value-based; on-policy; model-free; communication; discrete communication channel; discrete action space; continuous state space; cooperative task; centralized training with decentralized execution; multi-agent;

强化学习论文（1）MADDPG

less than 1 minute read

Published:

标签： MADDPG; actor-critic; off-policy; model-free; communication; continuous communication channel; continuous action space; continuous state space; mixed task; cooperative task; competitive task; centralized training with decentralized execution; multi-agent;

安装配置 supervisor 用于管理守护进程

1 minute read

Published:

Supervisor 是用 Python 开发的一套通用的进程管理程序，能将一个普通的命令行进程变为后台 daemon，并监控进程状态，异常退出时能自动重启。

强化学习思考（1）前言

less than 1 minute read

Published:

选择开设一个专栏来记录自己学习强化学习的过程，由于网上已经有很多基础强化学习的教程了，所以本专栏更多的关注点不在于基础，而是一些容易遗漏的点，一些细节上的补充，作为学习过程中的思考笔记。

pytorch 小技巧

2 minute read

Published:

记录一些写 pytorch 时候不会的点。

python 中配置 linux 环境变量

less than 1 minute read

Published:

在某些 python 程序中需要单独配置一些环境变量，又不希望通过 pycharm 配置，有时候比较麻烦。

服务器配置 jupyter notebook

1 minute read

Published:

希望实现在本地访问服务器上的 jupyter notebook / lab，目前远程使用 jupyter lab 加载较慢，而且插件还不够丰富，待以后再补充详细使用。

pycharm 配色方案

less than 1 minute read

Published:

在看莫烦老师的代码时感觉他的配色方案好好看，所以安装一个同款。

linux 小技巧

3 minute read

Published:

查看重要信息以及一些命令使用。

python 小技巧

5 minute read

Published:

开始写 python 之后才发现一堆不会的，有一点不会就查一点最后补充到这，未完待续。

Pandas 基础使用

12 minute read

Published:

Pandas 和 NumPy 有什么不同：如果用 python 的列表和字典来作比较, 那么可以说 Numpy 是列表形式的，没有数值标签，Pandas 就是字典形式，而且 Pandas 是基于 NumPy 构建的。

Windows 下搭建 FTP 服务器

less than 1 minute read

Published:

同时使用两台电脑的时候要互相传小文件用 u 盘太麻烦了，所以在 win 10 下搭建一个 FTP 服务器来共享文件。

python 文件开头两行注释解释

less than 1 minute read

Published:

常见 python 文件的前两行都会写上如下两行代码， 下面分别解释这两条语句的意义。

本地连接远程服务器 GUI 图形界面

less than 1 minute read

Published:

最近在使用服务器编程时遇到了无法实时显示可视化界面的问题，最多只能做到将图片保存下来，在本地下载之后再打开，下面我们将介绍 windows 和 mac 如何实现连接远程服务器跑 python 代码实时返回可视化界面的操作，当然配置成功后不局限于 python 运行结果的可视化，包括一切 linux GUI 图形界面都能做到，具体原理主要要参考 X 协议，在这篇博客有适当提及。

NumPy 基础使用

8 minute read

Published:

下面介绍关于 NumPy 的基本函数使用。

python 虚拟环境配置

1 minute read

Published:

先确保自己的用户目录下安装了 anaconda，接下来我们会使用 anaconda 来创建一个属于自己的虚拟环境，这样就不会被其他项目环境所干扰，无论在 mac 还是 linux 下都推荐使用虚拟环境。

pip 和 conda 包的转移复制

less than 1 minute read

Published:

有时候我们在一个地方已经配好了环境，希望能给另一个地方配置，就在当前环境下导出 requirements.txt，然后进入到另一个环境下通过 requirements.txt 安装所有的包。

conda 环境的转移复制

less than 1 minute read

Published:

有时候我们在一个地方已经配好了 conda 的虚拟环境，希望能直接复制给另一个地方，就在当前环境下导出 environment.yml，然后在另一台设备下通过 environment.yml 安装所有的包。

OMP: Error #15

less than 1 minute read

Published:

在 mac 环境下，运行机器学习代码时出现错误提示。

python 中利用 pickle 保存变量

less than 1 minute read

Published:

在编写 python 项目时，有时候希望将模型训练的结果进行保存，以便下次进行使用，在这里利用 pickle 实现。

使用 screen 后台运行命令避免 SSH 断连

less than 1 minute read

Published:

在使用服务器的过程中，通常会跑一些比较长时间的代码，这个时候希望能将程序运行在服务器后台，防止客户端断开 SSH 时结束程序。

使用 nohup 后台运行命令避免 SSH 断连

1 minute read

Published:

在使用服务器的过程中，通常会跑一些比较长时间的代码，这个时候希望能将程序运行在服务器后台，防止客户端断开 SSH 时结束程序。

pycharm 连接远程服务器

1 minute read

Published:

在写 python 项目时希望在本地编写代码，然后可以实时自动同步到远程服务器上进行运行，可以使用 pycharm 实现。

python 顺序执行多个 py 文件

less than 1 minute read

Published:

在做项目的过程中想写个脚本运行多个 py 文件，但是在本地运行 bat 文件还需要设置调用 pycharm 中设置的 python 虚拟环境，感觉比较麻烦，所以搜索了如果用 python 写代码来顺序执行多个 py 文件。

python 相对路径

less than 1 minute read

Published:

在 python 中读写文件需要用到各种相对路径符，下面进行简单说明。

LaTeX 手动安装宏包

less than 1 minute read

Published:

使用 CTeX 套装，需要使用一些没有预装的宏包，这时就需要自己安装 package 了，下面使用自己下载的 type1cm 宏包进行讲解。

在 matplotlib 中使用 LaTeX 渲染文本

less than 1 minute read

Published:

使用 CTeX 套装，需要使用一些没有预装的宏包，这时就需要自己安装package了，下面使用自己下载的 type1cm 宏包进行讲解。

python 中 import 用法

less than 1 minute read

Published:

如果是自己遍写的依赖包，又不想安装到 python 的相应目录，可以放到本目录里进行 import 进行调用。

anaconda 下安装 jupyter notebook

1 minute read

Published:

之前一直用不了 jupyter notebook 很愁，今晚终于搭建成功了，赶紧记录一下，并且还有一些对 anaconda 的理解及运用。

Github Pages 搭建博客

3 minute read

Published:

成功搭建出自己的博客，肯定是要记录一下搭建过程的啦。

Hello, World

less than 1 minute read

Published:

第一次搭建自己的博客，以程序员的尿性，理应喊出一句：“Hello, world”。

portfolio

publications

talks

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.