Views
No views yet
laion/CLIP-ViT-g-14-laion2B-s12B-b42K. The image encoder is finetuned with FARE at $\epsilon=2/255$.
To load this model use:1from transformers import CLIPProcessor, CLIPModel
2
3model_name = "LEAF-CLIP/OpenCLIP-ViT-g-FARE2"
4processor_name = "laion/CLIP-ViT-g-14-laion2B-s12B-b42K"
5
6model = CLIPModel.from_pretrained(model_name)
7processor = CLIPProcessor.from_pretrained(processor_name)