Downloads · 30 days
51
7% of all-time downloads
CLIP-ViT-L-rho50-k1-constrained-FARE2
LEAF-CLIP/CLIP-ViT-L-rho50-k1-constrained-FARE2
CLIP-ViT-L-rho50-k1-constrained-FARE2 is a feature extraction model from LEAF-CLIP. Use it when you need embeddings to search or compare text. It is set up for transformers. The card lists the license as mit.
[[Paper]](https://www.arxiv.org/abs/2506.03355) [[Code]](https://github.com/LIONS-EPFL/LEAF)
All-time downloads
748
Public
Parameters
428M
1.7 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.7 GB · 100%
At a glance
- Task
- Feature Extraction
- Library
- transformers
- License
- mit
- Model type
- clip
- Access
- Public
- Created
- Apr 16, 2025
- Updated
- Jun 8, 2025
- SHA
- f71d132a
Base models
Model card
From the Hugging Face model README
Model Initialized from openai/clip-vit-large-patch14. The image encoder is finetuned with FARE at $\epsilon=2/255$. The text encoder is finetuned with LEAF at $k=1$ with $\rho=50$ and semantic constraints.
To load this model use:
from transformers import CLIPProcessor, CLIPModel
model_name = "LEAF-CLIP/CLIP-ViT-L-rho50-k1-FARE2"
processor_name = "openai/clip-vit-large-patch14"
model = CLIPModel.from_pretrained(model_name)
processor = CLIPProcessor.from_pretrained(processor_name)
- Task
- Feature Extraction
- Library
- transformers
- Type
- clip
- License
- mit
- Created
- Apr 16, 2025
- Updated
- Jun 8, 2025