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,.,.) = 
#>   2.5568  -1.8705   2.5651  -2.8518  -0.4414
#>   -6.4169   7.6458   4.3671  -4.9875  -3.2369
#>   -3.1286   2.6792 -12.7062  -5.5724  -8.7022
#>    2.2325   1.8812  -3.8924  -0.1773  -2.4867
#>    3.2300  -3.1472  -5.2938  -5.6013   1.7491
#> 
#> (1,2,.,.) = 
#>   2.8212  -1.8808  -4.3737  -1.8916  -3.5754
#>   -1.9973   5.3080  -5.5791  -9.7201  -8.9990
#>   -2.0654  -6.1619  -0.6746 -11.2846  -9.5520
#>    1.7883  -6.1783  -2.7848   0.9098  -9.0128
#>   -1.1351  -3.0385  -6.0454   1.0130   5.4653
#> 
#> (1,3,.,.) = 
#>   3.4252  -4.1577  -1.6184  11.1418  -1.3712
#>    5.3313  -8.7328   2.1915   9.9304   4.7661
#>    7.5666  15.2094  -8.6994   3.0033   3.3304
#>   -7.6365  -7.9769  -0.5043   7.9092   4.4929
#>   -4.0548   3.7929  12.1501   2.7287  -2.3606
#> 
#> (1,4,.,.) = 
#>  -2.7799  -2.8323  14.1922  -6.5315   4.9636
#>    4.4233 -19.4052  -7.2880  -5.9969   7.4471
#>    5.2161  -7.7022   5.8766  -9.0684   7.8324
#>    5.4072  -3.8466  -7.1176  -6.3199   1.9464
#>  -10.2469   4.5499   4.1510  10.8789  -3.5523
#> 
#> (1,5,.,.) = 
#>   3.2937  -1.8881  11.0727   1.5660   1.9271
#> ... [the output was truncated (use n=-1 to disable)]
#> [ CPUFloatType{1,8,5,5} ]