Conv_transpose2d — torch_conv_transpose2d • torch

Conv_transpose2d

Usage

torch_conv_transpose2d(
  input,
  weight,
  bias = list(),
  stride = 1L,
  padding = 0L,
  output_padding = 0L,
  groups = 1L,
  dilation = 1L
)

Arguments

input: input tensor of shape \((\mbox{minibatch} , \mbox{in\_channels} , iH , iW)\)
weight: filters of shape \((\mbox{in\_channels} , \frac{\mbox{out\_channels}}{\mbox{groups}} , kH , kW)\)
bias: optional bias of shape \((\mbox{out\_channels})\). Default: NULL
stride: the stride of the convolving kernel. Can be a single number or a tuple (sH, sW). Default: 1
padding: dilation * (kernel_size - 1) - padding zero-padding will be added to both sides of each dimension in the input. Can be a single number or a tuple (padH, padW). Default: 0
output_padding: additional size added to one side of each dimension in the output shape. Can be a single number or a tuple (out_padH, out_padW). Default: 0
groups: split input into groups, \(\mbox{in\_channels}\) should be divisible by the number of groups. Default: 1
dilation: the spacing between kernel elements. Can be a single number or a tuple (dH, dW). Default: 1

conv_transpose2d(input, weight, bias=NULL, stride=1, padding=0, output_padding=0, groups=1, dilation=1) -> Tensor

Applies a 2D transposed convolution operator over an input image composed of several input planes, sometimes also called "deconvolution".

See nn_conv_transpose2d() for details and output shape.

Examples

if (torch_is_installed()) {

# With square kernels and equal stride
inputs = torch_randn(c(1, 4, 5, 5))
weights = torch_randn(c(4, 8, 3, 3))
nnf_conv_transpose2d(inputs, weights, padding=1)
}
#> torch_tensor
#> (1,1,.,.) = 
#>  -1.6616   0.0137  -6.1094  12.0189  -7.3687
#>    0.3018  -9.9261   5.7359   6.2925  -2.1599
#>   -6.8296   1.7846   0.4840  -0.1206  -1.0212
#>   -2.5626   2.5035  -2.3364   9.7845  -0.1807
#>   -4.9740   5.9314   1.8926   8.6423   0.5023
#> 
#> (1,2,.,.) = 
#>  -4.3267  -1.7866   1.5270  -3.7930  -2.4699
#>    7.9945  12.3674  -4.0553   5.6058   5.4137
#>   -1.1894  -9.2267  -3.7780   1.2540  -3.2867
#>    1.4207   4.9865  10.5490  -9.0030  -2.3086
#>   -1.1242  -2.8026  -3.6222  -0.3440  -0.8669
#> 
#> (1,3,.,.) = 
#>  -6.6568 -10.0631   4.5932   5.8927  -7.2034
#>    0.5252   7.9606  15.5193  -1.3555   9.6987
#>   -1.4041  -5.8996 -12.1034   3.6059  -0.2265
#>    6.3976  -6.2412   6.2445   5.7989   1.1949
#>    7.0675  -9.4311  -3.7631  -4.7934  -5.4831
#> 
#> (1,4,.,.) = 
#>  1.3087 -1.8244 -0.5475 -5.6405 -1.7390
#>   4.1300  0.9511 -3.7680 -8.6174  8.8853
#>   2.2635  4.2231 -8.2424 -5.2816  0.5352
#>  -1.5589 -1.6229  2.5225 -1.5586 -1.9212
#>  -0.8739 -1.6558 -2.4575 -4.6741  4.3333
#> 
#> (1,5,.,.) = 
#>  -8.7603   2.0489  -2.3902   4.7933  -1.4914
#> ... [the output was truncated (use n=-1 to disable)]
#> [ CPUFloatType{1,8,5,5} ]