densenet_264

lucid.models.densenet_264(num_classes: int = 1000, **kwargs) → DenseNet

The densenet_264 function constructs a DenseNet-264 model. This preset uses DenseNetConfig(block_config=[6, 12, 64, 48], growth_rate=32, num_init_features=64).

Total Parameters: 33,337,704

Function Signature

@register_model
def densenet_264(num_classes: int = 1000, **kwargs) -> DenseNet

Parameters

  • num_classes (int, optional): Number of output classes for the final fully connected layer. Default is 1000.

  • kwargs (dict, optional): Additional keyword arguments forwarded to DenseNetConfig, excluding the preset block_config, growth_rate, and num_init_features fields.

Returns

  • DenseNet: A DenseNet-264 model instance with the preset (block_config=[6, 12, 64, 48], growth_rate=32, num_init_features=64).