# torch.Tensor.is_shared¶

Tensor.is_shared()[source]

Checks if tensor is in shared memory.

This is always True for CUDA tensors.