Implementation of machine learning related papers and algorithms in (more or less) 101 lines of code.
Inspired by the great Papers-in-100-Lines-of-Code by Maxime Vandegar.
Flow Matching [code]
Flow Matching for Generative Modeling (Lipman et al., 2023)
https://openreview.net/forum?id=PqvMRDCJT9t
Free-form Flows [code]
Free-form Flows: Make Any Architecture a Normalizing Flow (Draxler et al., 2023)
https://arxiv.org/abs/2310.16624
Maximum Likelihood Training of Autoencoders (Sorrensonet al., 2023)
https://arxiv.org/abs/2306.01843