A lightweight implementation of argparse (python library for creating CLIs) in C

Reimplmentation of Online softmax in C++ and CUDA from the paper: https://arxiv.org/pdf/1805.02867

A minimal implementation of pytorch's autograd with full GPU compatibility. Taken inspiration from TinyGrad.

The agenda of this account.

Reimplementation of the paper: Faster Base64 Encoding and Decoding using AVX2 instructions (https://arxiv.org/abs/1704.00605)

Reimplementation of various tokenization algorithms in python and zig

Recreating BitLinear from the reference paper: https://arxiv.org/pdf/2310.11453