reference/generated/tensordict.TensorDictBase.clone

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

TensorDictBase.clone

TensorDictBase.clone(recurse: bool = True, **kwargs) → Self

Returns a copy of the tensordict, of the same type.

With recurse=True, every tensor leaf is copied. The result is not locked, even if the tensordict is: a clone is usually made to be modified. To get a TensorDict from any other type, call to_tensordict() instead.

Parameters:

• recurse (bool, optional) – if True, the leaves are copied too. If False, only the tree of nested tensordicts is copied: the nested tensordicts are new objects, and the leaves are the same tensors as in the tensordict. Defaults to True.

• **kwargs – options of the class, for example newfile, the destination file of a PersistentTensorDict cloned with recurse=True.

Examples

>>> import torch
>>> from tensordict import TensorDict
>>> td = TensorDict({"a": torch.zeros(3), "b": {"c": torch.ones(3)}}, [3])
>>> td_clone = td.clone()
>>> td_clone["a"] is td["a"]
False
>>> td_shallow = td.clone(recurse=False)
>>> td_shallow["a"] is td["a"], td_shallow["b"] is td["b"]
(True, False)
>>> td.lock_().clone().is_locked
False

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

Tutorials that use clone

Changing the shape of a TensorDict

Changing the shape of a TensorDict