torch.Tensor.is_pinned# Tensor.is_pinned()# Returns true if this tensor resides in pinned memory. By default, the device pinned memory on will be the current accelerator.