Webssl Dino3b Full2b 224

🚀 Web-SSL DINO ViT-3B: 2B MetaCLIP data, 224 Resolution

A 3 billion parameter Vision Transformer (ViT) trained with DINOv2 self - supervised learning on web - scale image data without language supervision.

🚀 Quick Start

Web-SSL DINO 3B is a powerful Vision Transformer model. It's trained via self - supervised learning on a large amount of web images without language supervision. This model can achieve comparable or better performance than language - supervised models in various vision tasks.

✨ Features

• High - Capacity Model: With 3 billion parameters, it has strong representational ability.
• Self - Supervised Learning: Trained on web - scale image data without language supervision, demonstrating the effectiveness of pure visual learning.
• Good Performance: Performs well on both traditional vision benchmarks and multimodal tasks such as visual question answering and OCR & chart understanding.

📦 Installation

The provided README doesn't contain specific installation steps, so this section is skipped.

💻 Usage Examples

Basic Usage

from transformers import AutoImageProcessor, Dinov2Model
import torch
from PIL import Image

processor = AutoImageProcessor.from_pretrained('facebook/webssl-dino3b-full2b-224')
# 'eager' and 'sdpa' attn_implementation supported
model = Dinov2Model.from_pretrained('facebook/webssl-dino3b-full2b-224')

# Process an image
image = Image.open('path/to/image.jpg')
inputs = processor(images=image, return_tensors="pt")
with torch.no_grad():
    outputs = model(**inputs)

cls_features = outputs.last_hidden_state[:, 0]  # CLS token features
patch_features = outputs.last_hidden_state[:, 1:] # patch-wise token features

📚 Documentation

Model Details

Model Descriptions

Web-SSL DINO 3B is a 3 billion parameter Vision Transformer model trained using self - supervised learning on 2 billion web images without language supervision. This model demonstrates that pure visual learning, when scaled appropriately, can match or exceed the performance of language - supervised models like CLIP across various vision tasks. It performs well on both traditional vision benchmarks and multimodal tasks including visual question answering and OCR & chart understanding.

📄 License

The model is licensed under cc - by - nc - 4.0.

📄 Citation

@article{fan2025scaling,
  title={Scaling Language-Free Visual Representation Learning}, 
  author={David Fan and Shengbang Tong and Jiachen Zhu and Koustuv Sinha and Zhuang Liu and Xinlei Chen and Michael Rabbat and Nicolas Ballas and Yann LeCun and Amir Bar and Saining Xie},
  year={2025},
  eprint={2504.01017},
  archivePrefix={arXiv},
  primaryClass={cs.CV}
}