[Question] Tracking status of of learning

Most contents of this QnA series are based on my own trouble, curiosity and sometimes for reminding.</br>

Reading list / Questions

Open Source

• The (written) unwritten guide to pull requests

  July 25, 2019 - https://www.atlassian.com/blog/git/written-unwritten-guide-pull-requests (and this blog has lots of good articles)

• Fastai contribution guide
• Fastai test guide ¹

Modern electronic circuit

• What is CUDA? 9.0, 10.0, 10.1 what differs? why torch conflicts when using not corresponded version?

• How to evaluate TPU spec with GPU? what criterion?

• How colab is limiting the GPU for colab pro user - here / so suppose I’m implementing a model and should I wait until virtually use GPU? / How awful the resource is if I use GPU all the time? ²

• High Performance Computing (Moore’s Law / Dennard Scaling)

Container (a.k.a. kubernetes/docker .. )

• Julia Evan’s article of container - https://jvns.ca/blog/2016/09/15/whats-up-with-containers-docker-and-rkt/

Github

• why clone using https (rather than ssh) is recommended?
• GIt Pro book
  • Why there is no changing remote from https to https / ssh <-> ssh? reference git pro document

Distributed / Parallel Computing

• CS224n Lecture2 prof. Christopher mentioned if we have lots of data(millions of word vectors), it is important to not have to send gigantic updates around
  • Don’t have any knowledge regarding this
  • (1) J. Han and B. Sharma, Learn CUDA Programming: A beginner’s guide to GPU programming and parallel computing with CUDA 10.x and C/C++, Packt, 2019.
  • (2) A. Sherif, and A. Ravindra, Apache Spark Deep Learning Cookbook: Over 80 recipes that streamline deep learning in a distributed environment with Apache Spark, Packt, 2018.
  • (3) B. Quinto, Next-Generation Machine Learning with Spark: Covers XGBoost, LightGBM, Spark NLP, Distributed Deep Learning with Keras, and More, apress, 2020
  • (4) G. Lozzia, Hands-On Deep Learning with Apache Spark: Build and deploy distributed deep learning applications on Apache Spark, Packt, 2019.

As a Researcher

• Normalized h5 index of CS conference

Pytorch

• what’s difference between DataLoader / DataLoaders?

  Data format

• Pickle: Usually saved with numpy format. is it required?
• ZIP, GZIP, 7z etc. many formats compress data
  • What does ‘entries’ mean in zip file?
  • when I zip list recursively (i.e. -r option) in command, I want to zip one parent containing contents (i.e. I’m in /usr and target archive files are only in/usr/root/sys/. How can I archive subdirectory of final path (i.e.sys/) without compressing other previous directory?

—-

1. wanted to run test to see specific example, reading document but couldn’t (crying) ↩

2. BTW, are GPUs at Casablanca? Why a timezone matches it? ↩