What's up guys,
Just wanted to share the humble beginnings of a Deep Learning framework I'm developing based on Sentdex's NNFS (Neural Networks From Scratch) book / YouTube series
I call it NimNet
The framework will be written in pure Nim- standard lib only, no dependencies. I had originally worked through parts 1-8 of the series using the neo library without issues, but I really wanted to understand the math involved behind everything... and a pure Nim Deep Learning framework just sounds really f*****g cool.
That being said, the math module I wrote is absolutely atrocious right now- completely unoptimized and lacks any exception handling. But it works for what's written so far! As I become a better Nim developer I'll be reworking the math and other implementations.
main.nim currently combines all of the functions & math I've written if you'd like to see what that looks like.
Sure is!
If you ever get into NNFS I do have parts 1-8 within nnfs directory you can refer to. I'm pretty sure there's a comment with the video reference as well