Aller au contenu
ModelPrésentéExterne
Télécharger

Oryx ViT

The Oryx-ViT model is trained on 200M data and can seamlessly and efficiently process visual inputs with arbitrary spatial sizes and temporal lengths. It is described in the paper Oryx MLLM: On-Demand...

Sourcehuggingface.co/THUdyh/Oryx-ViT/tree/657aa9c29f3b7edb41109b14a566015b67684f61
TypeOtherModèle de basegoogle/siglip-so400m-patch14-384VersionRevision 657aa9c

Source du modèle

Type
Other
Taille du fichier
852 MB
Modèle de base
google/siglip-so400m-patch14-384
Version
vRevision 657aa9c
Mis à jour
28 sept. 2026
Téléchargements
0
Source
Hugging Face

Description de la source

Oryx-ViT

Model Summary

The Oryx-ViT model is trained on 200M data and can seamlessly and efficiently process visual inputs with arbitrary spatial sizes and temporal lengths. It is described in the paper Oryx MLLM: On-Demand Spatial-Temporal Understanding at Arbitrary Resolution.

Model Architecture

  • Architecture: SigLip
  • Data: a mixture of 200M data, 2 epoch
  • Precision: BFloat16

Hardware & Software

  • Hardware: 64 * NVIDIA Tesla A100
  • Orchestration: HuggingFace Trainer
  • Code: Pytorch

Citation

@article{liu2024oryx,
title={Oryx MLLM: On-Demand Spatial-Temporal Understanding at Arbitrary Resolution},
author={Liu, Zuyan and Dong, Yuhao and Liu, Ziwei and Hu, Winston and Lu, Jiwen and Rao, Yongming},
journal={arXiv preprint arXiv:2409.12961},
year={2024}
}

Sources

1 source
Hugging FacePrincipalVersion 657aa9c29f3b7edb41109b14a566015b67684f61
huggingface.co/THUdyh/Oryx-ViT/tree/657aa9c29f3b7edb41109b14a566015b67684f61

Vérifié 28 sept.

Artefacts du modèle

1 artefact

oryx_vit.pth

pth · 852 MB · SHA-256 f7a4cdafb621…a6a7 · Hugging Face

Télécharger

Extraits de sources

2 extraits
Readme28 sept.huggingface.co/THUdyh/Oryx-ViT

--- base_model: - google/siglip-so400m-patch14-384 language: - en - zh license: apache-2.0 pipeline_tag: image-feature-extraction --- # Oryx-ViT ## Model Summary The Oryx-ViT model is trained on 200M data and can seamlessly and efficiently process visual inputs with arbitrary spatial sizes and temporal lengths. It is described in the paper [Oryx MLLM: On-Demand Spatial-Temporal Understanding at Arbitrary Resolution](https://arxiv.org/abs/2409.12961). - **Repository:** https://github.com/Oryx-mllm/Oryx - **Project Page:** https://oryx-mllm.github.io - **Languages:** English, Chinese ### Model Architecture - **Architecture:** SigLip - **Data:** a mixture of 200M data, 2 epoch - **Precision:** BFloat16 #### Hardware & Software - **Hardware:** 64 * NVIDIA Tesla A100 - **Orchestration:** HuggingFace Trainer - **Code:** Pytorch ## Citation ```bibtex @article{liu2024oryx, title={Oryx MLLM: On-Demand Spatial-Temporal Understanding at Arbitrary Resolution}, author={Liu, Zuyan and Dong, Yuhao and Liu, Ziwei and Hu, Winston and Lu, Jiwen and Rao, Yongming}, journal={arXiv preprint arXiv:2409.12961}, year={2024} } ```

Note28 sept.huggingface.co/THUdyh/Oryx-ViT

Source context: 0 downloads · 8 likes · Pipeline image-feature-extraction · Repo THUdyh/Oryx-ViT