tupone@gentoo.org Tupone Alfredo Support distributed applications Use FBGEMM Enable flash attention Use sci-libs/gloo Use sci-libs/mkl for blas, lapack and sparse blas routines Use NNPACK Add support for math operations through numpy Use oneDNN Use sci-libs/openblas for blas routines Use OpenMP for parallel code Use QNNPACK Enable ROCm gpu computing support Use XNNPACK pytorch/pytorch