代码库

Fast and Furious AMD Kernels
C++
FlashFFTConv: Efficient Convolutions for Long Sequences with Tensor Cores
C++
Tile primitives for speedy kernels
Cuda
Big & Small LLMs working together
Python