What is the explaination of each value when we print an Pytorch Model?

Question

I am working on a code downloaded from github. When i printed the pretrained model i get following in console output. Can anyone explain in graphical form that how each step can be shown in graphical form?

    DataParallel(
  (module): Unet(
    (encoder): EfficientNetEncoder(
      (conv_stem): Conv2d(4, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (act1): Swish()
      (blocks): Sequential(
        (0): Sequential(
          (0): DepthwiseSeparableConv(
            (conv_dw): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=64, bias=False)
            (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(64, 16, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(16, 64, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pw): Conv2d(64, 32, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn2): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Identity()
          )
          (1): DepthwiseSeparableConv(
            (conv_dw): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=32, bias=False)
            (bn1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(32, 8, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(8, 32, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pw): Conv2d(32, 32, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn2): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Identity()
          )
          (2): DepthwiseSeparableConv(
            (conv_dw): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=32, bias=False)
            (bn1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(32, 8, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(8, 32, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pw): Conv2d(32, 32, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn2): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Identity()
          )
          (3): DepthwiseSeparableConv(
            (conv_dw): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=32, bias=False)
            (bn1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(32, 8, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(8, 32, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pw): Conv2d(32, 32, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn2): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Identity()
          )
        )
        (1): Sequential(
          (0): InvertedResidual(
            (conv_pw): Conv2d(32, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(192, 192, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=192, bias=False)
            (bn2): BatchNorm2d(192, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(192, 8, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(8, 192, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(192, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): InvertedResidual(
            (conv_pw): Conv2d(48, 288, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(288, 288, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=288, bias=False)
            (bn2): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(288, 12, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(12, 288, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(288, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): InvertedResidual(
            (conv_pw): Conv2d(48, 288, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(288, 288, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=288, bias=False)
            (bn2): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(288, 12, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(12, 288, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(288, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): InvertedResidual(
            (conv_pw): Conv2d(48, 288, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(288, 288, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=288, bias=False)
            (bn2): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(288, 12, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(12, 288, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(288, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (4): InvertedResidual(
            (conv_pw): Conv2d(48, 288, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(288, 288, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=288, bias=False)
            (bn2): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(288, 12, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(12, 288, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(288, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (5): InvertedResidual(
            (conv_pw): Conv2d(48, 288, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(288, 288, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=288, bias=False)
            (bn2): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(288, 12, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(12, 288, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(288, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (6): InvertedResidual(
            (conv_pw): Conv2d(48, 288, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(288, 288, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=288, bias=False)
            (bn2): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(288, 12, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(12, 288, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(288, 48, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(48, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (2): Sequential(
          (0): InvertedResidual(
            (conv_pw): Conv2d(48, 288, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(288, 288, kernel_size=(5, 5), stride=(2, 2), padding=(2, 2), groups=288, bias=False)
            (bn2): BatchNorm2d(288, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(288, 12, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(12, 288, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(288, 80, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(80, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): InvertedResidual(
            (conv_pw): Conv2d(80, 480, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(480, 480, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=480, bias=False)
            (bn2): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(480, 20, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(20, 480, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(480, 80, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(80, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): InvertedResidual(
            (conv_pw): Conv2d(80, 480, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(480, 480, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=480, bias=False)
            (bn2): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(480, 20, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(20, 480, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(480, 80, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(80, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): InvertedResidual(
            (conv_pw): Conv2d(80, 480, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(480, 480, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=480, bias=False)
            (bn2): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(480, 20, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(20, 480, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(480, 80, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(80, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (4): InvertedResidual(
            (conv_pw): Conv2d(80, 480, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(480, 480, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=480, bias=False)
            (bn2): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(480, 20, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(20, 480, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(480, 80, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(80, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (5): InvertedResidual(
            (conv_pw): Conv2d(80, 480, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(480, 480, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=480, bias=False)
            (bn2): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(480, 20, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(20, 480, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(480, 80, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(80, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (6): InvertedResidual(
            (conv_pw): Conv2d(80, 480, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(480, 480, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=480, bias=False)
            (bn2): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(480, 20, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(20, 480, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(480, 80, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(80, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )
        (3): Sequential(
          (0): InvertedResidual(
            (conv_pw): Conv2d(80, 480, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(480, 480, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), groups=480, bias=False)
            (bn2): BatchNorm2d(480, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(480, 20, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(20, 480, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(480, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (1): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (2): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (3): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (4): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (5): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (6): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (7): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (8): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
          (9): InvertedResidual(
            (conv_pw): Conv2d(160, 960, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn1): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act1): Swish()
            (conv_dw): Conv2d(960, 960, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=960, bias=False)
            (bn2): BatchNorm2d(960, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (act2): Swish()
            (se): SqueezeExcite(
              (avg_pool): AdaptiveAvgPool2d(output_size=1)
              (conv_reduce): Conv2d(960, 40, kernel_size=(1, 1), stride=(1, 1))
              (act1): Swish()
              (conv_expand): Conv2d(40, 960, kernel_size=(1, 1), stride=(1, 1))
            )
            (conv_pwl): Conv2d(960, 160, kernel_size=(1, 1), stride=(1, 1), bias=False)
            (bn3): BatchNorm2d(160, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
          )
        )

i want output in the form as show in figure. How can i read this model summary and represent in the form of a diagram?

Noob ML Dude · Answer 1 · 2022-10-23T15:03:04.520

To visualize the network architechture in a graphical form, you could use Tensorboard. Tensorboard is a popular tool used for working with Neural Networks. It is useful for visualizing neural networks, visualizing datasets, plotting metric charts, etc. You can use Tensorboard even with PyTorch.

Based on your comment below, I understand you already have a neural network in your model variable that you are able to print. To visualize that neural network and its components you could install tensorboard by running the below command in a terminal.

pip install tensorboard

Once installed, you can run tensorboard using the below command in a terminal

tensorboard --logdir runs

Here runs/ is the directory in which tensorboard will look for logs. You don't need to create it, tensorboard will create it for you when you start writing below. When tensorboard starts, it will show where the URL where tensorboard server is running. You could now open this URL in your browser to see the Tensorboard application running.

Now you have tensorboard set up and running, you can use the tensorboard utilities in PyTorch (assuming you already have PyTorch installed) for working with Tensorboard. Using these utility functions you could add many objects (graphs, figures, etc) to Tensorboard. To see a visual representation of your neural network, you can run the minimal snippet below:

from torch.utils.tensorboard import SummaryWriter
writer = SummaryWriter('runs/unnet_experiment_1')

writer.add_graph(model, images) # model is the network class (using nn.Module)
writer.close()

This would add the neural network to Tensordboard, and you can view it on the Tensorboard UI under "Graphs" tab. It would looks similar to this: The boxes represent your neural network. You can double click on the boxes to show details of these components.

See this Pytorch Tensorboard Tutorial for more details on how to use Tensorboard with PyTorch.

For other alternatives to Tensorboard, see this SO answer which gives a few other package options for visualizing a PyTorch model.

thanks for your answer but i could not understand that. I have downloaded the code form github along with model weights folder(having model_best.pth files) and model logs folder(having config.yml files). Then i printed the model via following code. model = get_model(config) print(model) I want to visualize each step of the model to under stand what is going on in that. I am new to this please help me understand this in terms of a lay man. — Muhammad Salman, Oct 21 '22 at 11:12
Sure, I can explain in simpler ways. Could you tell me how much background you have for Deep Learning or PyTorch so I can tailor the answer accordingly. — Noob ML Dude, Oct 23 '22 at 14:23
i have just learn deep learning from few online courses on youtube. But i wanted to try a code for my thesis for Masters. I had run the code and got results but for paper writing i need to explain each step in graphical form. The code i used is from github with few changes in parameters. i just wanted to draw the model in graphical form. Unfortunately pytorch has no model.summary() function. So i just loaded the pre trained model in to a variable and just print it to get detail view of that. Now i want to show each step in graphical form can you please help me to draw that in graphical form? — Muhammad Salman, Oct 24 '22 at 11:29

What is the explaination of each value when we print an Pytorch Model?

1 Answers1