Model Card for vit_l14_pn_bioclip-v2

A ViT l14 image encoder from BioCLIP-2 by Gu et al., converted to the Birder format for image feature extraction. This version preserves the original model weights and architecture, with the exception of removing the CLIP projection layer to expose raw image embeddings. Trained on the large-scale TreeOfLife-200M dataset, it serves as a powerful foundation for downstream computer vision tasks. The model excels at understanding biological imagery across diverse taxonomic groups.

See: https://huggingface.co/imageomics/bioclip-2 for further details.

Model Details

• Model Type: Image classification and detection backbone

• Model Stats:

  • Params (M): 303.2
  • Input image size: 224 x 224

• Dataset: Trained on the TreeOfLife-200M dataset

• Papers:

  • An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale: https://arxiv.org/abs/2010.11929
  • BioCLIP 2: Emergent Properties from Scaling Hierarchical Contrastive Learning: https://arxiv.org/abs/2505.23883

Model Usage

Image Embeddings

import birder
from birder.inference.classification import infer_image

(net, model_info) = birder.load_pretrained_model("vit_l14_pn_bioclip-v2", inference=True)

# Get the image size the model was trained on
size = birder.get_size_from_signature(model_info.signature)

# Create an inference transform
transform = birder.classification_transform(size, model_info.rgb_stats)

image = "path/to/image.jpeg"  # or a PIL image
(out, embedding) = infer_image(net, image, transform, return_embedding=True)
# embedding is a NumPy array with shape of (1, 1024)

Detection Feature Map

from PIL import Image
import birder

(net, model_info) = birder.load_pretrained_model("vit_l14_pn_bioclip-v2", inference=True)

# Get the image size the model was trained on
size = birder.get_size_from_signature(model_info.signature)

# Create an inference transform
transform = birder.classification_transform(size, model_info.rgb_stats)

image = Image.open("path/to/image.jpeg")
features = net.detection_features(transform(image).unsqueeze(0))
# features is a dict (stage name -> torch.Tensor)
print([(k, v.size()) for k, v in features.items()])
# Output example:
# [('neck', torch.Size([1, 1024, 16, 16]))]

Citation

@misc{dosovitskiy2021imageworth16x16words,
      title={An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale},
      author={Alexey Dosovitskiy and Lucas Beyer and Alexander Kolesnikov and Dirk Weissenborn and Xiaohua Zhai and Thomas Unterthiner and Mostafa Dehghani and Matthias Minderer and Georg Heigold and Sylvain Gelly and Jakob Uszkoreit and Neil Houlsby},
      year={2021},
      eprint={2010.11929},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2010.11929},
}

@misc{gu2025bioclip2emergentproperties,
      title={BioCLIP 2: Emergent Properties from Scaling Hierarchical Contrastive Learning},
      author={Jianyang Gu and Samuel Stevens and Elizabeth G Campolongo and Matthew J Thompson and Net Zhang and Jiaman Wu and Andrei Kopanev and Zheda Mai and Alexander E. White and James Balhoff and Wasila Dahdul and Daniel Rubenstein and Hilmar Lapp and Tanya Berger-Wolf and Wei-Lun Chao and Yu Su},
      year={2025},
      eprint={2505.23883},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2505.23883},
}

@software{ilharco_gabriel_2021_5143773,
  author={Ilharco, Gabriel and Wortsman, Mitchell and Wightman, Ross and Gordon, Cade and Carlini, Nicholas and Taori, Rohan and Dave, Achal and Shankar, Vaishaal and Namkoong, Hongseok and Miller, John and Hajishirzi, Hannaneh and Farhadi, Ali and Schmidt, Ludwig},
  title={OpenCLIP},
  year={2021},
  doi={10.5281/zenodo.5143773},
}

Acknowledgments

This model is based on the excellent work by Gu et al. in BioCLIP-2, who developed the training methodology, curated the TreeOfLife-200M dataset, and trained this powerful biological image understanding model. The implementation builds upon the OpenCLIP framework by Ilharco et al., which made this scale of contrastive learning possible. All credit for the model's capabilities goes to these original authors. This conversion simply adapts their work to the Birder framework format.