wide_resnet_101¶
The wide_resnet_101 function creates a Wide ResNet-101 architecture, a ResNet-101 variant with wider bottleneck layers for increased representational capacity.
Total Parameters: 126,886,696
Function Signature¶
def wide_resnet_101(num_classes: int = 1000, **kwargs) -> ResNet
Parameters¶
num_classes (int, optional): The number of output classes for the final fully connected layer. Default is 1000.
kwargs (dict, optional): Additional keyword arguments forwarded to ResNetConfig, excluding the preset block and layers fields. Any provided block_args are merged with the default {“base_width”: 128} wide bottleneck setting.
Returns¶
ResNet: A Wide ResNet-101 model configured with the specified number of output classes and additional options.
Architecture Details¶
Wide ResNet-101 uses the preset ResNetConfig(block=”bottleneck”, layers=[3, 4, 23, 3], block_args={“base_width”: 128}).
Note
The architecture comprises 4 stages with layers [3, 4, 23, 3] and wider bottleneck blocks.
Examples¶
Creating a Wide ResNet-101 model for ImageNet classification:
import lucid.models as models
model = models.wide_resnet_101(num_classes=1000)
print(model)