Pretrained weights for lucid.models.efficientformer_l3_cls.
Single ImageNet-1k checkpoint converted from timm's
efficientformer_l3.snap_dist_in1k (Li et al., 2022; ~31.4M params,
82.4% top-1).
Examples
>>> from lucid.models.vision.efficientformer._weights import EfficientFormerL3Weights
>>> list(EfficientFormerL3Weights.__members__)
['SNAP_DIST_IN1K', 'DEFAULT']
DEFAULT is an alias rather than a fourth entry, so a bare
pretrained=True and the tag it resolves to cannot drift apart.
>>> EfficientFormerL3Weights.DEFAULT is EfficientFormerL3Weights.SNAP_DIST_IN1K
True
>>> EfficientFormerL3Weights.SNAP_DIST_IN1K.num_classes
1000