microsoft / ai-system Goto Github PK
View Code? Open in Web Editor NEWSystem for AI Education Resource.
Home Page: https://microsoft.github.io/AI-System/
License: Creative Commons Attribution 4.0 International
ai-system's Issues
textbook 8.3
- 8.3.5 是否可以考虑放到8.7去，叫做“8.7 - 实验”
- 图 8.3.6 看起来很模糊，可能需要重绘
- 参考文献 7625 字，太长
README.md中，关于“A-基础教程”的超链接失效
README.md中，关于“A-基础教程”的超链接失效。
看起来应该是：
https://github.com/microsoft/ai-edu/tree/master/%E5%9F%BA%E7%A1%80%E6%95%99%E7%A8%8B
textbook 11
- 11.1 模型压缩，能不能实现一个简单的算法，把一个现成的模型压缩后看看前后比对的inference效果 ? 如果training过程比较难的话就写个Inference的。数据量化，稀疏化，知识蒸馏，轻量化，张量分解，随便选一个当例子。
- 不要有四级目录 a.b.c.d，最多只能三级，用 xxxxx 表示四级。
- 图11-1-1这种，点和文字重合在一起，不仔细看都看不出来，图上也没有网格。需要保证类似图的质量。
textbook 3 and 4
- 个人感觉第3章和第4章可以合并（从内容连贯性上看）。另外一个理由就是这两章内容都比较少。
basiclab lab1 mnist_tensorboard.py have some problems
my environment : CUDAtoolkit 10.0 pytorch 1.5.0 tensorflow 1.15.0
when i run python mnist_tensorboard.py , some error happened:
2021-03-21 20:15:01.978418: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_100.dll
2021-03-21 20:15:04.781186: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_100.dll
Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same
Error occurs, No graph saved
Traceback (most recent call last):
File "mnist_tensorboard.py", line 199, in <module>
main()
File "mnist_tensorboard.py", line 182, in main
writer.add_graph(model, images)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\utils\tensorboard\writer.py", line 707, in add_graph
self._get_file_writer().add_graph(graph(model, input_to_model, verbose))
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\utils\tensorboard\_pytorch_graph.py", line 291, in graph
raise e
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\utils\tensorboard\_pytorch_graph.py", line 285, in graph
trace = torch.jit.trace(model, args)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\jit\__init__.py", line 875, in trace
check_tolerance, _force_outplace, _module_class)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\jit\__init__.py", line 1027, in trace_module
module._c._create_method_from_trace(method_name, func, example_inputs, var_lookup_fn, _force_outplace)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\nn\modules\module.py", line 548, in __call__
result = self._slow_forward(*input, **kwargs)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\nn\modules\module.py", line 534, in _slow_forward
result = self.forward(*input, **kwargs)
File "mnist_tensorboard.py", line 61, in forward
x = self.conv1(x)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\nn\modules\module.py", line 548, in __call__
result = self._slow_forward(*input, **kwargs)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\nn\modules\module.py", line 534, in _slow_forward
result = self.forward(*input, **kwargs)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\nn\modules\conv.py", line 349, in forward
return self._conv_forward(input, self.weight)
File "D:\Program_Files\Anaconda3\envs\ai-system-learn\lib\site-packages\torch\nn\modules\conv.py", line 346, in _conv_forward
self.padding, self.dilation, self.groups)
RuntimeError: Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same
I think there are something wrong with tensorboard. How can I fix this?
openpai 1.2 报错”No worker node is detected.“
在执行bash quick-start-service.sh -m ~/master.csv -w ~/worker.csv -c ~/config.yaml报错”No worker node is detected.“，查询发现错误在执行/contrib/kubespray/script/openpai-generator.py (line 304)脚本时出现。请问是否有解决方法。
Issues in chapter 6
请移除保持章节一致性
textbook 10
- 以图10.1.1举例，这里没有a,s,r的符号，读者无法和后面的解释对应起来
- 10.1.2 的Gt公式是可以有展开形式的，如果写出来会更容易理解
-
$\gamma$ 我记得是 (0,1], 不是 [0,1) -
$\pi (a|s)=p(a_t=a|s_t=s)$ 这类的公式（我认为的）标准写法是$\pi (a|s)=p(A_t=a|S_t=s)$ - 10.1 参考文献内容太长，与正文不成比例
- 10.1.2 可能需要一个三个圆叠加的图来展示三者的关系
- 10.1.1 和 10.1.2 请放在一个md文件中
- 10.2.2 中为什么会有四级索引序号的图？比如 图10.2.2.1，可以改成 10.2.1
- 10.2.2 和 10.2.3 可以考虑合并为 10.3，这样篇幅和内容上都可以和 10.2.1（变成10.2）来匹配
Micr2
Issues in chapter 12
1
2 Please add all the concepts (中文（英文）) in 术语表.md
制作了教材的PDF
请问教材是还在写作当中吗？
您好！我看见Textbook目录下缺少一些章节的内容。请问本课程的教材是还在写作当中吗？以及本课程是否有配套的视频可以观看呢？
Content in Security
Dear Xian and Peichen,
我们可以在安全章节添加总览内容并展开到各章吗？例如课程中这两张图分别放入内容。统揽全局
lab 6,7,9
应该是对应到哪个章节的呢？能否合并到正文中？
title update ? 安全与隐私章节
- 原则 一致性和简洁
服务-> 推理，减少问题词出现次数
2 标题修改
原始
12. 人工智能安全与隐私
12.1 人工智能内在安全与隐私
12.1.1 深度神经网络的内在安全问题
12.1.2 深度神经网络的内在隐私问题
12.2 人工智能训练安全与隐私
12.2.1 深度学习训练时的安全问题
12.2.2 深度学习训练时的隐私问题
12.2.3 联邦学习及其训练时的隐私问题
12.3 人工智能服务安全与隐私
12.3.1 深度学习服务时的安全问题
12.3.2 深度学习服务时的用户隐私问题
12.3.3 深度学习服务时的模型隐私问题
->
新
12. 人工智能安全与隐私
12.1 人工智能内在安全与隐私
12.1.1 深度神经网络的安全问题
12.1.2 深度神经网络的隐私问题
12.2 人工智能训练安全与隐私
12.2.1 训练系统安全
12.2.2 训练系统隐私
12.2.3 联邦学习隐私
12.3 人工智能推理安全与隐私
12.3.1 推理系统安全
12.3.2 推理系统用户隐私
12.3.3 推理系统隐私
textbook 8.1
正文6000字，参考文献4000字，比例失衡，缩短后者
术语格式与引用链接 （in Chapter 安全与隐私）
Hello Xian，请参考术语格式将英文首字母大写，引用工作 由于markdown比较方便加链接，可以都加上可以点进去的链接哈。
例如
textbook 1.1
markdown文件
- Line 27~36，第二层缩进后的，没有并列子内容，就可以取消前面的 bullet，比如 “- 谷歌、百度、..." -> "\t谷歌、百度、..."
- Line49~58，不能即用 bullet 又用 序号 (1)(2)(3)
- 图序号的标准写法是 ”图 1.1.1“
- 1.1.3，神经网络的基本理论在深度学习前已基本奠定，有点儿啰嗦，可以是”神经网络基本理论的奠定“
- Line 70 - 90, 可以用 bullet缩进
- line 92, 逗号改为顿号
questions on lab6
Dear teachers, we have problems doing our homework-lab6, we are now stuck in one step, the problem is shown in the picture, could you please help us?thanks a lot！
Make RL to chapter 10 and add readme link
textbook 6
- 表6-5-1，横向不够的话改成纵向，否则无法印刷
- 表6-3-1，表头不需要用 ** 加黑，它自黑。该表格的内容最好居中或者右对齐。
- 并行算法，有对应的 lab 吗？没有的话是否可以简单实现一个来加以说明？
Issues in chapter 5
1
2 英文大小写似乎和前面不一致
textbook 1.2
- line 20, 开头加回车
- 表 1-2-1，标准写法是 表 1.2.1，并且应该放在表的上方
- line 96-106，二级缩进时，不需要bullet
- line 115, 10的8次方倍，直接写成
$10^8$ 倍 - 图 1.2.3 中的右上侧的蜻蜓、老鼠、人脑，没有解释其含义。V100 上有个 3？
- line 131,137, 为何与 line 57,75,90 不一致？
部署OpenPai遇到问题
老师，您好。我们在部署openpai的过程的第二步部署部署 Kubernetes的过程中，运行sudo bash quick-start-kubespray.sh -m ~/master.csv -w ~/worker.csv -c ~/config.yaml后master与worker机器出现了源冲突的问题，如下图。
请教您该如何解决？
add GammarRegressor
usually, only the training change, so it is just about using the same converter for a new model.
textbook 12
- 图变形了，不用设置 width, height
- 12.1 差分隐私SGD算法，可不可以给个具体的数据例子来写一段代码跑一下？
- 12.2.1 数据隐私保护，能不能做一个简单的prototype，写代码来实现并说明一下？
- 12.2.2 模型隐私保护，能不能实现一个简单的水印技术来说明？
mistakes in textbook 3.1-基于数据流图的深度学习框架
这里表述有误 tensorflow应该是声明式 Chainer和PyTroch是命令式 这里写反了
这里举例的文字和给出的示例不匹配
或许我可以开个PR来订正吗？
video
Great job! Will there be an open video course?
textbook 8.4
- 参考文献，太长
关于教材视频链接
想问下该教材有没有上课的视频资料链接呢？
Issues in chapter 11
Hi Shijie,
Repo中我们不维护slides，请删除，并放入teams folders
Issues in chapter 13
参考章节
An Error by Lecture 3 - Computation frameworks for DNN
In page 19 by calculation of the gradient of L(x), there is probably an error in the second part for sin(exp(x)+exp(x)^2
Should the gradient be cos(exp(x)+exp(x)^2)(exp(x)+2exp(x)^2), i.e. there is one unnecessary exp in the answer?
Thanks
video recording of the course
Thanks for your great work. Is there an open video recording of the course? If so, could you share the link?
openpai k8s error
Starting kubernetes...
setup k8s cluster
PLAY [localhost] *******************************************************************************************************************************************************************************************
[WARNING]: Could not match supplied host pattern, ignoring: bastion
PLAY [bastion[0]] ******************************************************************************************************************************************************************************************
skipping: no hosts matched
PLAY [k8s-cluster:etcd] ************************************************************************************************************************************************************************************
included: /home/openpai/pai-deploy/kubespray/roles/bootstrap-os/tasks/bootstrap-debian.yml for stu-276, iair279, stu-282
PLAY [k8s-cluster:etcd] ************************************************************************************************************************************************************************************
TASK [kubernetes/preinstall : Stop if access_ip is not pingable] *******************************************************************************************************************************************
changed: [iair279]
changed: [stu-276]
changed: [stu-282]
included: /home/openpai/pai-deploy/kubespray/roles/container-engine/docker/tasks/set_facts_dns.yml for stu-276, iair279, stu-282
[WARNING]: flush_handlers task does not support when conditional
TASK [download : prep_download | Create staging directory on remote node] **********************************************************************************************************************************
changed: [stu-276]
changed: [iair279]
changed: [stu-282]
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/prep_kubeadm_images.yml for stu-276
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_file.yml for stu-276
TASK [download : download_file | Create dest directory on node] ********************************************************************************************************************************************
changed: [stu-276]
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/extract_file.yml for stu-276
[WARNING]: noop task does not support when conditional
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_file.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_file.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_file.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_file.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/download_container.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/set_docker_image_facts.yml for stu-276
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/check_pull_required.yml for stu-276
TASK [download : download_file | Create dest directory on node] ********************************************************************************************************************************************
changed: [stu-282]
changed: [iair279]
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/extract_file.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/extract_file.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/extract_file.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/extract_file.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/set_docker_image_facts.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/check_pull_required.yml for stu-276, iair279, stu-282
TASK [download : download_container | Download image if required] ******************************************************************************************************************************************
changed: [stu-276 -> 192.168.1.187]
changed: [iair279 -> 192.168.1.187]
changed: [stu-282 -> 192.168.1.187]
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/set_docker_image_facts.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/check_pull_required.yml for stu-276, iair279, stu-282
TASK [download : download_container | Download image if required] ******************************************************************************************************************************************
changed: [stu-276 -> 192.168.1.187]
changed: [stu-282 -> 192.168.1.187]
changed: [iair279 -> 192.168.1.187]
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/set_docker_image_facts.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/check_pull_required.yml for stu-276, iair279, stu-282
TASK [download : download_container | Download image if required] ******************************************************************************************************************************************
changed: [stu-276 -> 192.168.1.187]
changed: [iair279 -> 192.168.1.187]
changed: [stu-282 -> 192.168.1.187]
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/set_docker_image_facts.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/check_pull_required.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/set_docker_image_facts.yml for stu-276, iair279, stu-282
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/check_pull_required.yml for stu-276, iair279, stu-282
TASK [download : download_container | Download image if required] ******************************************************************************************************************************************
changed: [iair279 -> 192.168.1.187]
changed: [stu-276 -> 192.168.1.187]
changed: [stu-282 -> 192.168.1.187]
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/set_docker_image_facts.yml for stu-276
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/check_pull_required.yml for stu-276
TASK [download : download_container | Download image if required] ******************************************************************************************************************************************
changed: [stu-276 -> 192.168.1.187]
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/set_docker_image_facts.yml for stu-276
included: /home/openpai/pai-deploy/kubespray/roles/download/tasks/check_pull_required.yml for stu-276
FAILED - RETRYING: download_container | Download image if required (4 retries left).
FAILED - RETRYING: download_container | Download image if required (3 retries left).
FAILED - RETRYING: download_container | Download image if required (2 retries left).
FAILED - RETRYING: download_container | Download image if required (1 retries left).
TASK [download : download_container | Download image if required] ******************************************************************************************************************************************
fatal: [stu-276 -> 192.168.1.187]: FAILED! => {"attempts": 4, "changed": true, "cmd": ["/usr/bin/docker", "pull", "k8s.gcr.io/cluster-proportional-autoscaler-amd64:1.6.0"], "delta": "0:00:15.027078", "end": "2021-05-11 19:29:07.932611", "msg": "non-zero return code", "rc": 1, "start": "2021-05-11 19:28:52.905533", "stderr": "Error response from daemon: Get https://k8s.gcr.io/v2/: net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)", "stderr_lines": ["Error response from daemon: Get https://k8s.gcr.io/v2/: net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)"], "stdout": "", "stdout_lines": []}
NO MORE HOSTS LEFT *****************************************************************************************************************************************************************************************
PLAY RECAP *************************************************************************************************************************************************************************************************
iair279 : ok=204 changed=7 unreachable=0 failed=0 skipped=192 rescued=0 ignored=0
localhost : ok=1 changed=0 unreachable=0 failed=0 skipped=0 rescued=0 ignored=0
stu-276 : ok=295 changed=8 unreachable=0 failed=1 skipped=258 rescued=0 ignored=0
stu-282 : ok=204 changed=7 unreachable=0 failed=0 skipped=192 rescued=0 ignored=0
Issues in chapter 10
Hi Hui , 请增加类似摘要的一些问题，吸引读者。例如，RL的引出，Recall哪些传统问题，重要性motivation之类的words。Please reference chapter 13
部署 OpenPAI
老师，我在执行这条指令的时候出现了这样的错误，请问一下该怎样解决？
Issues in Chapter 2
MSRA-USTC Practice Project Discussion
Welcome to discuss about the practice projects here!
Lack part of contents at chapter 2
Dear Xiaowu哥,
Seems Chapter 2 lacks some content below. Will add it in the later version?
2.2 深度学习系统基础
2.2.1 深度学习运算的表示
2.2.2 编译框架与中间表达
2.2.3 运行态和硬件
2.2.4 分布式执行
2.2.5 深度学习系统性能优化
figure is too small please remove width setting
figure broken link in 12.3
New contents 补充in chapter 2
Dear Xiaowu哥，当前transformer, MOE and Pathways系列模型越来越重要，可否帮补一章这类模型结构？与CNN和RNN并列
https://github.com/microsoft/AI-System/tree/main/Textbook/%E7%AC%AC2%E7%AB%A0-%E7%A5%9E%E7%BB%8F%E7%BD%91%E7%BB%9C%E5%9F%BA%E7%A1%80
textbook 9
- 自动化机器学习，是我们做过的NNI吗？我看到了 lab 8, 将来这个lab 8 会印到书里吗？如果不会，建议把lab 8的内容添加到第9章中以丰满之。
textbook 1.4
textbook 10
- 章节序号错误
textbook 13.2
- 13.2.4, 每行包括不多的字，没有形成整体段落，很难理解作者要表达什么。
- 13.2.3 GP, MPI, EI, UCB, 都是很好的展开知识点。即使假设读者都知道这些概念（我觉得可能性不大），也应该结合具体应用场景来讲解一下细节。书就是这么写的。
2021_USTC_JointPhD - 人工智能系统实践 - 成果提交
本 issue 将作为2021_USTC_JointPhD-人工智能系统实践项目成果的提交地址
项目截止日期： 5月26日
请在北京时间2021年5月27日17点前，在本issue下提交你的成果。
提交步骤：
- 创建公共仓库
请各位同学使用个人的Github账户创建名为MSRA-USTC-AISystemProject-2021的公开仓库 - 上传成果
请各位同学按以下目录结构，将个人成果进行整理，删去个人隐私信息，并上传至第一步中创建的公开仓库中
# 实验1
|-- lab 1
# 图片均放在images目录下
|-- images
|-- image1.png
|-- image2.jpg
# 代码均放在src目录下
|-- src
|-- code1.py
|-- code2.ipynb
# 其他文件均放在resources目录下
|-- resources
# 请在README.md中放置各个文件的内容说明，你的实验流程和成果展示
|-- README.md
# 实验6
|-- lab 6
|-- ...
- 提交成果
按以下内容格式回复本issue，请注意保护个人隐私
1. 报名时提供的邮件地址: yourmail[at]your.domain
2. 个人仓库地址: https://github.com/yourname/yourrepositories
3. 补充信息(可选)有其他疑问的地方请随时与助教联系
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.