Expand description
CPU/GPU kernel implementations of teenygrad’s nn layers
(nn — including a Flash Attention 2 forward/backward implementation, math ops, and
graph lowering), written against the teeny-triton DSL and compiled via teeny-compiler.
The cuda feature (on by default) enables the teeny-cuda backend, which requires the CUDA
toolkit to build — see its crate README. Running/compiling kernels additionally needs the
custom teenyc compiler at runtime; see teeny-compiler’s README.