nvidia.dali.fn.squeeze#

nvidia.dali.fn.squeeze(__data, /, *, axes=None, axis_names=None, bytes_per_sample_hint=[0], preserve=False, device=None, name=None)#

Removes the dimensions given as axes or axis_names.

It’s an error to remove a dimension that would cause the total volume to change.

This operator allows sequence inputs and supports volumetric data.

Supported backends

‘cpu’
‘gpu’

Parameters:

__data¶ (TensorList) – Data to be squeezed

Keyword Arguments:

axes¶ (int or list of int or TensorList of int, optional) –
Indices of dimensions which should be removed.

All squeezed dimensions should have size 1, unless the total volume of the tensor is 0 before and after squeeze. All indices must be in the range of valid dimensions of the input
axis_names¶ (layout str, optional) –
Layout columns which should be removed.

All squeezed dimensions should have size 1, unless the total volume of the tensor is 0 before and after squeeze. All layout names should be present in data layout.
bytes_per_sample_hint¶ (int or list of int, optional, default = [0]) –
Output size hint, in bytes per sample.

If specified, the operator’s outputs residing in GPU or page-locked host memory will be preallocated to accommodate a batch of samples of this size.
preserve¶ (bool, optional, default = False) – Prevents the operator from being removed from the graph even if its outputs are not used.