cvt_13

lucid.models.cvt_13(num_classes: int = 1000, **kwargs) → CvT

The cvt_13 function constructs the CvT-13 preset. This preset uses the default CvTConfig stage layout for the 13-layer variant.

Total Parameters: 19,997,480

Function Signature

@register_model
def cvt_13(num_classes: int = 1000, **kwargs) -> CvT

Parameters

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

  • kwargs (dict, optional): Additional keyword arguments forwarded to CvTConfig, excluding the preset num_stages, patch_size, patch_stride, patch_padding, dim_embed, num_heads, and depth fields.

Returns

  • CvT: A CvT model instance constructed from the CvT-13 preset config.