A Python-embedded DSL that makes it easy to write fast, scalable ML kernels with minimal boilerplate.
Open on GitHub