Upsample¶

class Upsample(target_size, mode='bilinear')[source]¶

Bases: sphinx.ext.autodoc.importer._MockObject

Module to perform Upsampling to a given size. Should replace the deprecated torch.nn.Upsample2d

[summary]

Parameters:	target_size (int or tuple) – output spatial size. mode (str) – algorithm used for upsampling: ‘nearest’ \| ‘linear’ \| ‘bilinear’ \| ‘trilinear’ \| ‘area’. Default: ‘nearest’

forward(input_tensor)[source]¶

Performs the actual upsampling

Parameters:	input_tensor (`torch.Tensor`) – the input tensor to be upsampled
Returns:	upsampled tensor
Return type:	`torch.Tensor`