timm / ViT-L-16-SigLIP-384

huggingface.co
Total runs: 140.2K
24-hour runs: 0
7-day runs: -87.4K
30-day runs: -298.7K
Model's Last Updated: October 25 2023
zero-shot-image-classification

Introduction of ViT-L-16-SigLIP-384

Model Details of ViT-L-16-SigLIP-384

Model card for ViT-L-16-SigLIP-384

A SigLIP (Sigmoid loss for Language-Image Pre-training) model trained on WebLI.

This model has been converted to PyTorch from the original JAX checkpoints in Big Vision . These weights are usable in both OpenCLIP (image + text) and timm (image only).

Model Details
Model 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-L-16-SigLIP-384')
tokenizer = get_tokenizer('hf-hub:timm/ViT-L-16-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_large_patch16_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
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}}
}

Runs of timm ViT-L-16-SigLIP-384 on huggingface.co

140.2K
Total runs
0
24-hour runs
-29.8K
3-day runs
-87.4K
7-day runs
-298.7K
30-day runs

More Information About ViT-L-16-SigLIP-384 huggingface.co Model

More ViT-L-16-SigLIP-384 license Visit here:

https://choosealicense.com/licenses/apache-2.0

ViT-L-16-SigLIP-384 huggingface.co

ViT-L-16-SigLIP-384 huggingface.co is an AI model on huggingface.co that provides ViT-L-16-SigLIP-384's model effect (), which can be used instantly with this timm ViT-L-16-SigLIP-384 model. huggingface.co supports a free trial of the ViT-L-16-SigLIP-384 model, and also provides paid use of the ViT-L-16-SigLIP-384. Support call ViT-L-16-SigLIP-384 model through api, including Node.js, Python, http.

ViT-L-16-SigLIP-384 huggingface.co Url

https://huggingface.co/timm/ViT-L-16-SigLIP-384

timm ViT-L-16-SigLIP-384 online free

ViT-L-16-SigLIP-384 huggingface.co is an online trial and call api platform, which integrates ViT-L-16-SigLIP-384's modeling effects, including api services, and provides a free online trial of ViT-L-16-SigLIP-384, you can try ViT-L-16-SigLIP-384 online for free by clicking the link below.

timm ViT-L-16-SigLIP-384 online free url in huggingface.co:

https://huggingface.co/timm/ViT-L-16-SigLIP-384

ViT-L-16-SigLIP-384 install

ViT-L-16-SigLIP-384 is an open source model from GitHub that offers a free installation service, and any user can find ViT-L-16-SigLIP-384 on GitHub to install. At the same time, huggingface.co provides the effect of ViT-L-16-SigLIP-384 install, users can directly use ViT-L-16-SigLIP-384 installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

ViT-L-16-SigLIP-384 install url in huggingface.co:

https://huggingface.co/timm/ViT-L-16-SigLIP-384

Url of ViT-L-16-SigLIP-384

ViT-L-16-SigLIP-384 huggingface.co Url

Provider of ViT-L-16-SigLIP-384 huggingface.co

timm
ORGANIZATIONS

Other API from timm

huggingface.co

Total runs: 19.0M
Run Growth: -1.8M
Growth Rate: -9.67%
Updated: January 21 2025
huggingface.co

Total runs: 4.9M
Run Growth: 3.3M
Growth Rate: 68.60%
Updated: January 21 2025
huggingface.co

Total runs: 125.5K
Run Growth: 103.6K
Growth Rate: 82.60%
Updated: October 25 2023
huggingface.co

Total runs: 121.9K
Run Growth: -120.0K
Growth Rate: -98.99%
Updated: January 21 2025
huggingface.co

Total runs: 113.9K
Run Growth: -112.3K
Growth Rate: -100.13%
Updated: January 21 2025
huggingface.co

Total runs: 112.5K
Run Growth: 101.7K
Growth Rate: 89.79%
Updated: January 21 2025
huggingface.co

Total runs: 90.3K
Run Growth: 64.4K
Growth Rate: 80.38%
Updated: January 21 2025
huggingface.co

Total runs: 31.4K
Run Growth: 8.7K
Growth Rate: 27.99%
Updated: January 21 2025
huggingface.co

Total runs: 24.0K
Run Growth: 7.5K
Growth Rate: 31.41%
Updated: January 21 2025
huggingface.co

Total runs: 22.4K
Run Growth: 15.1K
Growth Rate: 70.96%
Updated: January 21 2025
huggingface.co

Total runs: 20.8K
Run Growth: 6.5K
Growth Rate: 32.11%
Updated: January 21 2025
huggingface.co

Total runs: 20.5K
Run Growth: 6.0K
Growth Rate: 30.02%
Updated: January 21 2025
huggingface.co

Total runs: 20.5K
Run Growth: 5.8K
Growth Rate: 28.79%
Updated: January 21 2025
huggingface.co

Total runs: 20.2K
Run Growth: 15.2K
Growth Rate: 76.29%
Updated: January 21 2025