reference/generated/tensordict.TensorDictBase.is_shared

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

TensorDictBase.is_shared

TensorDictBase.is_shared() → bool

Whether the tensordict is in shared memory.

It is True after share_memory_(), and always for a tensordict on a CUDA device. A tensordict built from tensors that are in shared memory is not marked as shared: only share_memory_() marks it. A tensordict in shared memory is locked: entries cannot be added, renamed or removed. See Locking and temporary changes.

Examples

>>> import torch
>>> from tensordict import TensorDict
>>> td = TensorDict(a=torch.zeros(3), batch_size=[3])
>>> td.is_shared()
False
>>> td.share_memory_().is_shared()
True
>>> td.is_locked
True

Every tensordict class and every tensorclass has this method. For the other methods, see Methods by task.