General PyTorch Questions

Hi everyone,

I am interested in contributing to PyTorch and had a couple of questions about the community:

  1. I was wondering about your perspective on how approachable this project is to someone who does not have experience with this software, given that I am a senior in college?

  2. Do you have any idea on the tasks that I would be able to contribute to? Are there good first issues for someone with not a lot of experience with machine learning?

Hey,

Welcome!

  1. It might not be the easiest project to dive in mostly because it has grown mostly organically and some parts are not very well documented. But you can ask questions either on the related issue or slack and people will be happy to help you discover the different part of the code and guide you with what needs to be modified. But in general, the codebase goes from lots of high level python code (including the domain libraries) that is very ML focused to a lot of low level code to handle fast operations and the many features pytorch provides.
  2. I would recommend to take a look at https://github.com/pytorch/pytorch/blob/master/CONTRIBUTING.md. For good first issue, it depends a lot of what kind of problem you’re interested in. High level ML? Python typing? c++/python binding? jit? distributed? low level kernel implementations? etc