convnext_v2_huge

lucid.models.convnext_v2_huge(num_classes: int = 1000, **kwargs) → ConvNeXt_V2

The convnext_v2_huge function constructs the ConvNeXt-v2 Huge preset. This preset uses the default ConvNeXtV2Config stage layout for the huge variant.

Total Parameters: 660,289,640

Function Signature

@register_model
def convnext_v2_huge(num_classes: int = 1000, **kwargs) -> ConvNeXt_V2

Parameters

  • num_classes (int, optional): Number of output classes for classification. Default is 1000.

  • kwargs (dict, optional): Additional keyword arguments forwarded to ConvNeXtV2Config, excluding the preset depths and dims fields.

Returns

  • ConvNeXt_V2: A ConvNeXt-v2 model instance constructed from the huge preset config.