xma.functional.fused_residual_add_rmsnorm.triton_implementation.forward

fused_residual_add_rmsnorm_forward_triton_kernel(x_ptr, x_stride, r_ptr, r_stride, W_ptr, W_stride, y_ptr, y_stride, xr_ptr, xr_stride, s_ptr, s_stride, eps, multiplier, B, H, BLOCK_SIZE_B: triton.language.constexpr, BLOCK_SIZE_H: triton.language.constexpr)