ml4gw.nn package

Subpackages

• ml4gw.nn.autoencoder package
  • Submodules
  • ml4gw.nn.autoencoder.base module
    • Autoencoder
      • Autoencoder.decode()
      • Autoencoder.encode()
      • Autoencoder.forward()
  • ml4gw.nn.autoencoder.convolutional module
    • ConvBlock
      • ConvBlock.decode()
      • ConvBlock.encode()
    • ConvolutionalAutoencoder
      • ConvolutionalAutoencoder.decode()
      • ConvolutionalAutoencoder.forward()
  • ml4gw.nn.autoencoder.skip_connection module
    • AddSkipConnect
      • AddSkipConnect.forward()
    • ConcatSkipConnect
      • ConcatSkipConnect.forward()
      • ConcatSkipConnect.get_out_channels()
    • SkipConnection
      • SkipConnection.forward()
      • SkipConnection.get_out_channels()
  • ml4gw.nn.autoencoder.utils module
    • match_size()
  • Module contents
• ml4gw.nn.resnet package
  • Submodules
  • ml4gw.nn.resnet.resnet_1d module
    • BasicBlock
      • BasicBlock.expansion
      • BasicBlock.forward()
    • Bottleneck
      • Bottleneck.expansion
      • Bottleneck.forward()
    • BottleneckResNet1D
      • BottleneckResNet1D.block
    • ResNet1D
      • ResNet1D.block
      • ResNet1D.forward()
    • conv1()
    • convN()
  • ml4gw.nn.resnet.resnet_2d module
    • BasicBlock
      • BasicBlock.expansion
      • BasicBlock.forward()
    • Bottleneck
      • Bottleneck.expansion
      • Bottleneck.forward()
    • BottleneckResNet2D
      • BottleneckResNet2D.block
    • ResNet2D
      • ResNet2D.block
      • ResNet2D.forward()
    • conv1()
    • convN()
  • Module contents
• ml4gw.nn.streaming package
  • Submodules
  • ml4gw.nn.streaming.online_average module
    • OnlineAverager
      • OnlineAverager.forward()
      • OnlineAverager.get_initial_state()
  • ml4gw.nn.streaming.snapshotter module
    • Snapshotter
      • Snapshotter.forward()
      • Snapshotter.get_initial_state()
  • Module contents

Submodules

ml4gw.nn.norm module

class ml4gw.nn.norm.GroupNorm1D(num_channels, num_groups=None, eps=1e-05)

Bases: Module

Custom implementation of GroupNorm which is faster than the out-of-the-box PyTorch version at inference time.

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses. :rtype: Float[Tensor, 'batch channel length']

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

class ml4gw.nn.norm.GroupNorm1DGetter(groups=None)

Bases: object

Utility for making a NormLayer Callable that maps from an integer number of channels to a torch Module. Useful for command-line parameterization with jsonargparse.

class ml4gw.nn.norm.GroupNorm2DGetter(groups=None)

Bases: object

Utility for making a NormLayer Callable that maps from an integer number of channels to a torch Module. Useful for command-line parameterization with jsonargparse.

Module contents