Register activation kernel layers on XPU (#47858)
kernels-community/activation now ships an XPU backend, so map the six layers
it provides to XPU as well. They were previously registered for CUDA only,
which left XPU on the eager fallback.
FastGELU, NewGELU and QuickGELU have no native PyTorch operator and expand
into several element-wise launches in eager mode, so the fused kernel is
several times faster. SiLU, GeLU and GeluTanh are on par with eager and are
mapped for parity with CUDA.
Co-authored-by: regisss <15324346+regisss@users.noreply.github.com>