API for Models

class libcll.models.Linear(input_dim, num_classes=10)

Bases: Module

Parameters:

• input_dim (int) – the feature space after data compressed into a 1D dimension.

• num_classes (int) – the number of classes.

training: bool

class libcll.models.MLP(input_dim, hidden_dim, num_classes=10)

Bases: Module

Parameters:

• input_dim (int) – the feature space after data compressed into a 1D dimension.

• hidden_dim (int) – the hidden dimension.

• num_classes (int) – the number of classes.

training: bool

libcll.models.ResNet

alias of <module ‘libcll.models.ResNet’ from ‘/home/docs/checkouts/readthedocs.org/user_builds/libcll/envs/latest/lib/python3.8/site-packages/libcll/models/ResNet.py’>

class libcll.models.DenseNet(num_layers=[16, 16, 16], dropRate=0, num_classes=10, growthRate=12, compressionRate=2)

Bases: Module

Parameters:

• num_layers (list) – the number of layers of each dense block.

• dropRate (int) – the drop rate of the dropout layers.

• growthRate (int) – the additional number of channels for each layer.

• compressionRate (int) – the ratio by which the number of feature maps is reduced at transition layers.

• num_classes (int) – the number of classes.

training: bool