Pretrained weights for lucid.models.clip_vit_base_32.
Trained on WIT-400M, the paper's 400-million image-text pair
dataset. num_classes is 0 because CLIPModel has no fixed label set —
that is the point of it.
Examples
>>> from lucid.models.multimodal.clip._weights import CLIPViTBase32Weights
>>> list(CLIPViTBase32Weights.__members__)
['OPENAI_WIT400M', 'DEFAULT']
DEFAULT is an alias rather than a fourth entry, so a bare
pretrained=True and the tag it resolves to cannot drift apart.
>>> CLIPViTBase32Weights.DEFAULT is CLIPViTBase32Weights.OPENAI_WIT400M
True
>>> CLIPViTBase32Weights.OPENAI_WIT400M.num_classes
0