Hello HN community - for the past year or so I have been plugging away on a side project called torchutils (https://gitlab.com/avilay/torchutils). It'd be great if you could take it for a spin when training your next ML model and give me some feedback. Here are the detailed docs (https://avilay.gitlab.io/torchutils/)
It would be helpful if you provided some context to why you developed these utils, like what exactly were you doing, what were the pain points, why haven't you adopted any of the existing tools, pros and cons of your approach, etc.