0%

问题描述:

在build docker的时候出现报错

1
The repository 'https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64  InRelease' is not signed.

仓库如下

1
pytorch/pytorch:1.10.0-cuda11.3-cudnn8-devel

解决方案:

https://tickets.dominodatalab.com/hc/en-us/articles/6101281885076-Build-error-for-nvidia-GPG-key

A new key is available and can be installed in your build with the following process. Take note of the inline comments as they are important.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
#fix nvidia gpg key
# if you also have gpg/pgsql/sbt fix, put this before
RUN \
#The following files may not exist in your environment and the commands will fail the build if they don't,
#so comment out the next two lines as needed
sudo rm /etc/apt/sources.list.d/cuda.list && \
sudo rm /etc/apt/sources.list.d/nvidia-ml.list && \
#
#The ubuntu version in the URL below is important, note this example is for Ubuntu 18.04.
#You can find your ubuntu version with the command in a workspace terminal.
#lsb_release -a
#
#Update the URL as per the following nvidia article
#https://developer.nvidia.com/blog/updating-the-cuda-linux-gpg-repository-key/
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/cuda-keyring_1.0-1_all.deb && \
sudo apt-key del 7fa2af80 && \
sudo dpkg -i cuda-keyring_1.0-1_all.deb

NOTE: when adding the additional instructions make sure the new RUN commands are executed prior to any other update operations (apt or yum), otherwise the run will fail .

导入数据

1
2
data = pd.read_csv('something.csv') #导入csv
df = pd.DataFrame(data) #构建DataFrame

操作数据

1
2
3
4
5
6
df.loc[df["class"] == 1] #针对某列进行筛选
df.loc["class"] #取出指定列
df.iloc[:2,:3] #取出前两行,前三列
df.shape #[row,col]
pd.concat(a,b) #合并
positive = positive.sample(frac=1) #shuffle frac是随机取样的百分比

可能会问的问题

  • 计算机网络ISO七层架构是什么
  • img

技术栈

  • 蒙特卡洛搜索树
  • Minimax 算法
  • alpha-beta 剪枝

思路

为什么要写博客

之前自己并没有写笔记或者博客日记之类的习惯,偶尔在CSDN上写一两篇,但是在网络上看了很多别人写的博客或者学习笔记,对自己帮助很大,所以打算开始写自己的博客。其实写博客也是对自己的一种激励吧,只有能够将问题讲解给别人才能算真正学会了。

Hexo

Hexo有很多大佬都在推荐,据说很好用,但是我之前并没有什么感觉,甚至感觉静态文件的部署方式很奇怪,会不会很麻烦,于是今天来体验一把。

上手之后给我的感觉还是蛮惊艳的,我的方法是在服务器端创建一个git裸仓库,然后用hooks在每次push之后更新静态文件,只需要执行hexo g -d就可以完成更新,还是挺方便的,而且NextT真的很好看( ̄▽ ̄*)!这个主题灵活度还是挺高的,可以自定义一些东西,很喜欢。