vp_suite.model_blocks.conv

This module contains convolutional model blocks.

class DCGANConv(in_channels, out_channels, stride)

Bases: vp_suite.base.base_model_block.VPModelBlock

The class implements a DCGAN conv layer, as introduced in Radford et al. (arxiv.org/abs/1511.06434).

NAME: str = 'DCGAN - Conv'

The clear-text name for this model block.

PAPER_REFERENCE = 'arxiv.org/abs/1511.06434'

The publication where this model was introduced first.

__init__(in_channels, out_channels, stride)

Initializes internal Module state, shared by both nn.Module and ScriptModule.

forward(x)

Defines the computation performed at every call.

Should be overridden by all subclasses.

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.

training: bool

class DCGANConvTranspose(in_channels, out_channels, stride)

Bases: vp_suite.base.base_model_block.VPModelBlock

The class implements a DCGAN convTranspose layer, as introduced in Radford et al. (arxiv.org/abs/1511.06434).

NAME: str = 'DCGAN - ConvTranspose'

The clear-text name for this model block.

PAPER_REFERENCE = 'arxiv.org/abs/1511.06434'

The publication where this model was introduced first.

__init__(in_channels, out_channels, stride)

Initializes internal Module state, shared by both nn.Module and ScriptModule.

forward(x)

Defines the computation performed at every call.

Should be overridden by all subclasses.

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.

training: bool

class DoubleConv2d(in_channels, out_channels)

Bases: vp_suite.base.base_model_block.VPModelBlock

This class implements a 2D double-conv block, as used in the popular UNet architecture (Ronneberger et al., arxiv.org/abs/1505.04597).

NAME: str = 'DoubleConv2d'

The clear-text name for this model block.

PAPER_REFERENCE = 'arxiv.org/abs/1505.04597'

The publication where this model was introduced first.

__init__(in_channels, out_channels)

Initializes internal Module state, shared by both nn.Module and ScriptModule.

forward(x)

Defines the computation performed at every call.

Should be overridden by all subclasses.

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.

training: bool

class DoubleConv3d(in_channels, out_channels)

Bases: vp_suite.base.base_model_block.VPModelBlock

The class implements a 3D double-conv block, an extension of the DoubleConv2d block to also process the time dimension.

NAME: str = 'DoubleConv3d'

The clear-text name for this model block.

__init__(in_channels, out_channels)

Initializes internal Module state, shared by both nn.Module and ScriptModule.

forward(x)

Defines the computation performed at every call.

Should be overridden by all subclasses.

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.

training: bool