This repository performs similar GEMM optimization that was written by SIBOEHM. However, instead do it manually by writting all the kernels, it uses the abstraction layer and action space in tinygrad.
optimize matrix multiplication using tinygrad's action space
This repository performs similar GEMM optimization that was written by SIBOEHM. However, instead do it manually by writting all the kernels, it uses the abstraction layer and action space in tinygrad.
optimize matrix multiplication using tinygrad's action space