Last released Jul 11, 2026
A declarative, ultra-minimalist ML framework for zero-boilerplate hardware-agnostic inference and training.
Supported by
