I had some free time so I have been upgrading KlongPy.
KlongPy now supports PyTorch as a backend and autograd over the array language.
A while back someone on HN thought it would be cool to have Klong support autograd and I also thought that would be cool.
There's a lot of examples in https://github.com/briangu/klongpy/tree/main/examples/autogr...
but here's a taste:
f::{x^2} :" Define f(x) = x^2 f:>3 :" Compute f'(3) = 6.0
Looking for feedback on new operators and any obvious gaps. Thanks!
updated github pages: https://klongpy.org repo: https://github.com/briangu/klongpy