resnest_50_4s2x40d

lucid.models.resnest_50_4s2x40d(num_classes: int = 1000, **kwargs) ResNeSt

The resnest_50_4s2x40d function creates an instance of the ResNeSt-50 variant configured with specific hyperparameters (4s2x40d), which enhance its representational capabilities for image recognition tasks.

Total Parameters: 30,417,464

Function Signature

@register_model
def resnest_50_4s2x40d(num_classes: int = 1000, **kwargs) -> ResNeSt

Parameters

  • num_classes (int, optional): Number of output classes for the classification task. Defaults to 1000.

  • kwargs (dict, optional): Additional keyword arguments passed to the ResNeSt constructor, allowing further customization of the model.

Returns

  • ResNeSt: An instance of the ResNeSt-50 model, configured with the provided parameters and 4s2x40d settings.

Hyperparameter Configuration

  • 4s: Indicates 4 splits in the Split-Attention mechanism (radix = 4), allowing the model to compute attention weights over 4 groups.

  • 2x: Specifies a cardinality of 2, meaning each group processes a subset of channels with two separate convolution operations.

  • 40d: Denotes a base width of 40, which scales the number of channels for intermediate feature maps.

Examples

from lucid.models import resnest_50_4s2x40d

# Create a ResNeSt-50 model for 10-class classification
model = resnest_50_4s2x40d(num_classes=10)

# Forward pass with a sample input
input_tensor = lucid.random.randn((1, 3, 224, 224))
output = model(input_tensor)
print(output.shape)  # Output: (1, 10)