Vit SO400M 14 SigLIP 384

🚀 ViT-SO400M-14-SigLIP-384

A SigLIP (Sigmoid loss for Language-Image Pre-training) model trained on WebLI, enabling zero-shot image classification.

This SigLIP model, trained on the WebLI dataset, offers a powerful solution for zero-shot image classification. It has been converted from the original JAX checkpoints in Big Vision to PyTorch, making it compatible with both OpenCLIP for combined image and text processing and timm for image-only tasks.

🚀 Quick Start

This SigLIP model, trained on the WebLI dataset, is a powerful tool for zero-shot image classification. It has been converted from the original JAX checkpoints in Big Vision to PyTorch, and can be used with both OpenCLIP (for image + text) and timm (for image only).

✨ Features

• Contrastive Image-Text: Capable of learning relationships between images and text.
• Zero-Shot Image Classification: Can classify images without prior training on specific classes.

📦 Installation

Ensure you have the necessary libraries installed:

pip install open-clip-torch>=2.23.0 timm>=0.9.8 torch pillow

💻 Usage Examples

Basic Usage

With OpenCLIP

import torch
import torch.nn.functional as F
from urllib.request import urlopen
from PIL import Image
from open_clip import create_model_from_pretrained, get_tokenizer # works on open-clip-torch>=2.23.0, timm>=0.9.8

model, preprocess = create_model_from_pretrained('hf-hub:timm/ViT-SO400M-14-SigLIP-384')
tokenizer = get_tokenizer('hf-hub:timm/ViT-SO400M-14-SigLIP-384')

image = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
image = preprocess(image).unsqueeze(0)

labels_list = ["a dog", "a cat", "a donut", "a beignet"]
text = tokenizer(labels_list, context_length=model.context_length)

with torch.no_grad(), torch.cuda.amp.autocast():
    image_features = model.encode_image(image)
    text_features = model.encode_text(text)
    image_features = F.normalize(image_features, dim=-1)
    text_features = F.normalize(text_features, dim=-1)

    text_probs = torch.sigmoid(image_features @ text_features.T * model.logit_scale.exp() + model.logit_bias)

zipped_list = list(zip(labels_list, [round(p.item(), 3) for p in text_probs[0]]))
print("Label probabilities: ", zipped_list)

With timm (for image embeddings)

from urllib.request import urlopen
from PIL import Image
import timm

image = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))

model = timm.create_model(
    'vit_so400m_patch14_siglip_384',
    pretrained=True,
    num_classes=0,
)
model = model.eval()

# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)

output = model(transforms(image).unsqueeze(0))  # output is (batch_size, num_features) shaped tensor

📚 Documentation

Model Details

📄 License

This model is licensed under the Apache-2.0 license.

📚 Citation

@article{zhai2023sigmoid,
  title={Sigmoid loss for language image pre-training},
  author={Zhai, Xiaohua and Mustafa, Basil and Kolesnikov, Alexander and Beyer, Lucas},
  journal={arXiv preprint arXiv:2303.15343},
  year={2023}
}

@misc{big_vision,
  author = {Beyer, Lucas and Zhai, Xiaohua and Kolesnikov, Alexander},
  title = {Big Vision},
  year = {2022},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/google-research/big_vision}}
}