---
title: metaclip-b16-400m
canonical_url: "https://www.modelscope.cn/models/facebook/metaclip-b16-400m"
md_url: "https://www.modelscope.cn/models/facebook/metaclip-b16-400m.md"
repository: facebook/metaclip-b16-400m
last_updated: 2025-05-20
license: cc-by-nc-4.0
model_type:
  - clip
architectures:
  - CLIPModel
library_name:
  - pytorch
frameworks:
  - pytorch
inference_backends:
  - "deploy_task vlm"
  - "sglang 0.5.2"
downloads: 345
stars: 0
tags:
  - vision
  - metaclip
---

# metaclip-b16-400m

> metaclip-b16-400m - facebook 在 ModelScope 开源的模型。MetaCLIP model, base-sized version, patch resolution 16

facebook/metaclip-b16-400m 是 ModelScope 魔搭社区上的机器学习模型，采用 cc-by-nc-4.0 许可，可用 deploy_task vlm、sglang 0.5.2 部署。

- **Repository**: facebook/metaclip-b16-400m
- **License**: cc-by-nc-4.0
- **Inference backends**: deploy_task vlm, sglang 0.5.2
- **Tags**: vision, metaclip
- **Downloads**: 345
- **Stars**: 0
- **Last updated**: 2025-05-20

Source: https://www.modelscope.cn/models/facebook/metaclip-b16-400m

---

# MetaCLIP model, base-sized version, patch resolution 16

MetaCLIP model applied to 400 million data points of CommonCrawl (CC). It was introduced in the paper [Demystifying CLIP Data](https://arxiv.org/abs/2309.16671) by Xu et al. and first released in [this repository](https://github.com/facebookresearch/MetaCLIP). 

Disclaimer: The team releasing MetaCLIP did not write a model card for this model so this model card has been written by the Hugging Face team.

## Model description

The [Demystifying CLIP Data](https://arxiv.org/abs/2309.16671) paper aims to reveal CLIP’s method around training data curation. OpenAI never open-sourced code regarding their data preparation pipeline.

<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/clip_overview.jpg"
alt="drawing" width="600"/>

<small> CLIP high-level overview. Taken from the <a href="https://arxiv.org/abs/2103.00020">CLIP paper</a>. </small>

## Intended uses & limitations

You can use the raw model for linking images with text in a shared embedding space. This enables things like zero-shot image classification, text-based image retrieval, image-based text retrieval, etc.

### How to use

We refer to the [docs](https://huggingface.co/docs/transformers/main/en/model_doc/clip#usage). Just replace the names of the models on the hub.

### BibTeX entry and citation info

```bibtex
@misc{xu2023demystifying,
      title={Demystifying CLIP Data}, 
      author={Hu Xu and Saining Xie and Xiaoqing Ellen Tan and Po-Yao Huang and Russell Howes and Vasu Sharma and Shang-Wen Li and Gargi Ghosh and Luke Zettlemoyer and Christoph Feichtenhofer},
      year={2023},
      eprint={2309.16671},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}
```