crested.tl.zoo.utils.conv_block

Contents

crested.tl.zoo.utils.conv_block#

crested.tl.zoo.utils.conv_block(inputs, filters, kernel_size, pool_size, activation, conv_bias=True, dropout=0.1, normalization='batch', res=False, padding='valid', l2=1e-05, batchnorm_momentum=0.99, name_prefix=None)#

Convolution building block.

Parameters:

  • inputs (KerasTensor) – Input tensor representing the data.

  • filters (int) – Number of filters in the convolutional layer.

  • kernel_size (int) – Size of the convolutional kernel.

  • pool_size (int) – Size of the max-pooling kernel.

  • activation (str) – Activation function applied after convolution.

  • conv_bias (bool (default: True)) – Whether to use bias in the convolutional layer (default is True).

  • dropout (float (default: 0.1)) – Dropout rate (default is 0.1).

  • normalization (str (default: 'batch')) – Type of normalization (‘batch’ or ‘layer’).

  • res (bool (default: False)) – Whether to use residual connections (default is False).

  • padding (str (default: 'valid')) – Padding type for the convolutional layer (default is “valid”).

  • l2 (float (default: 1e-05)) – L2 regularization weight (default is 1e-5).

  • batchnorm_momentum (float (default: 0.99)) – Batch normalization momentum (default is 0.99).

  • name_prefix (str | None (default: None)) – Prefix for layer names.

Return type:

KerasTensor

Returns:

The output tensor of the convolution block.

Contents