Hi everyone,
I’m Julian, an ML engineer with experience in distributed systems, and I’m currently setting up my local development environment to contribute to PyTorch.
I’m particularly interested in the torch.distributed.pipelining module, and I noticed it’s evolving from PiPPy. I would love to get involved in contributing to it. I have already forked and cloned the repo, and I’m currently inside a Dev Container building everything from source. I’m familiar with model parallelism concepts and would be happy to help with code, testing, or documentation tasks in this area.
If there are any beginner-friendly issues or areas where help is needed in the pipeline parallelism module, I’d really appreciate some guidance on where to start.
Thanks in advance, and I’m looking forward to contributing!