Class: TorchVision::Models::BasicBlock

Inherits:
Torch::NN::Module
  • Object
show all
Defined in:
lib/torchvision/models/basic_block.rb

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(inplanes, planes, stride: 1, downsample: nil, groups: 1, base_width: 64, dilation: 1, norm_layer: nil) ⇒ BasicBlock

Returns a new instance of BasicBlock.



4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
 
# File 'lib/torchvision/models/basic_block.rb', line 4

def initialize(inplanes, planes, stride: 1, downsample: nil, groups: 1, base_width: 64, dilation: 1, norm_layer: nil)
  super()
  norm_layer ||= Torch::NN::BatchNorm2d
  if groups != 1 || base_width != 64
    raise ArgumentError, "BasicBlock only supports groups=1 and base_width=64"
  end
  if dilation > 1
    raise NotImplementedError, "Dilation > 1 not supported in BasicBlock"
  end
  # Both self.conv1 and self.downsample layers downsample the input when stride != 1
  @conv1 = Torch::NN::Conv2d.new(inplanes, planes, 3, stride: stride, padding: 1, groups: 1, bias: false, dilation: 1)
  @bn1 = norm_layer.new(planes)
  @relu = Torch::NN::ReLU.new(inplace: true)
  @conv2 = Torch::NN::Conv2d.new(planes, planes, 3, stride: 1, padding: 1, groups: 1, bias: false, dilation: 1)
  @bn2 = norm_layer.new(planes)
  @downsample = downsample
  @stride = stride
end

Class Method Details

.expansion ⇒ Object 



41
42
43
 
# File 'lib/torchvision/models/basic_block.rb', line 41

def self.expansion
  1
end

Instance Method Details

#forward(x) ⇒ Object 



23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
 
# File 'lib/torchvision/models/basic_block.rb', line 23

def forward(x)
  identity = x

  out = @conv1.call(x)
  out = @bn1.call(out)
  out = @relu.call(out)

  out = @conv2.call(out)
  out = @bn2.call(out)

  identity = @downsample.call(x) if @downsample

  out += identity
  out = @relu.call(out)

  out
end