reference/generated/tensordict.TensorDictBase.grad

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

TensorDictBase.grad

property TensorDictBase.grad: Any

A tensordict of the .grad of every leaf, or None if no leaf has one.

Leaves whose .grad is None are left out. Assigning td.grad = other sets the .grad of each leaf from the matching leaf of other, or copies into it if it exists.

Examples

>>> import torch
>>> from tensordict import TensorDict
>>> td = TensorDict(x=torch.ones(3, requires_grad=True), y=torch.ones(3, requires_grad=True), batch_size=[3])
>>> (td["x"] * 2).sum().backward()
>>> td.grad["x"], "y" in td.grad.keys()
(tensor([2., 2., 2.]), False)

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