One more ML framework will fix all problems

Which is why I decided to start making one :) Working on Tensorlib just to brush up my C++.

Original inspiration for it was tinygrad. While tinygrad started off as an ambitious project - write good software for any accelerator to run - I feel it has devolved a little into an obfuscation contest to reduce the number of lines in its codebase. Still a great project though, just feels a little unapproachable to me.

I want tensorlib to be an educational project, for me or anyone reading the codebase. Will work on that after I finally manage to add two tensors.