reference/generated/tensordict.TensorDictBase.repeat

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

TensorDictBase.repeat

TensorDictBase.repeat(*repeats: int, inplace: bool = False) → Self

Repeats the batch along each batch dimension, as torch.Tensor.repeat() does.

The whole batch is repeated, as with numpy.tile(). To repeat each element next to itself instead, as with numpy.repeat(), use repeat_interleave(). Unlike expand(), it copies the data. The feature dimensions are not repeated, and the names are kept.

Parameters:

*repeats (int) – the number of repetitions along each batch dimension, one per batch dimension, as integers or as one torch.Size, list, tuple or tensor.

Keyword Arguments:

inplace (bool, optional) – if True, the tensordict keeps its identity and keys, its leaves are replaced one by one, and the call returns self. A lazy stack, or a tensordict that holds one, raises a NotImplementedError: call to_tensordict() first. See In-place operations. Defaults to False.

Raises:

ValueError – if the number of repeats differs from the number of batch dimensions.

Examples

>>> import torch
>>> from tensordict import TensorDict
>>> td = TensorDict(a=torch.tensor([[0, 1]]), b=torch.zeros(1, 2, 3), batch_size=[1, 2])
>>> out = td.repeat(2, 2)
>>> out["a"]
tensor([[0, 1, 0, 1],
        [0, 1, 0, 1]])
>>> out["b"].shape
torch.Size([2, 4, 3])

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