Things I’ve been doing (also check out my Github profile):

Chronos. Pretrained time series forecasting models, based on transformer architectures for language modeling. Some like to call these “time series foundation models”, but I do not.

ProtoGrad. An experimental Julia package for gradient-based optimization of machine learning models. Essentially, a highly opinionated collection of design ideas of mine, on how deep learning frameworks should work.

GluonTS. Python toolkit for probabilistic time series modeling, with a focus on deep learning architectures, built around PyTorch.

ProximalAlgorithms.jl. Generic Julia implementations of first-order optimization algorithms for nonsmooth problems, based on operator splittings: forward-backward (proximal gradient method), Douglas-Rachford (ADMM), primal-dual, and Davis-Yin splitting algorithms. Also contains Newton-type extensions. Based on:

ProximalOperators.jl. Julia package to compute the proximal operator of several functions commonly used in nonsmooth optimization problems. Useful as building block to implement large-scale optimization algorithms such as ADMM.

podds. Multi-threaded Texas hold ’em poker odds evaluation tool, written in C, command line only.

matto. Simple chess player implemented in C. I started this when I was 17 and learning the C programming language, so there’s a lot of room for improvement. Yet it plays!