README.md

---
dataset_info:
  features:
  - name: text
    dtype: string
  - name: images
    sequence: binary
  splits:
  - name: train
    num_bytes: 94020556918
    num_examples: 1465964
  download_size: 73033984223
  dataset_size: 94020556918
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---
# MMEB train split used in MoCa Continual Pre-training

[🏠 Homepage](https://haon-chen.github.io/MoCa/) | [💻 Code](https://github.com/haon-chen/MoCa) | [🤖 MoCa-Qwen25VL-7B](https://huggingface.co/moca-embed/MoCa-Qwen25VL-7B) | [🤖 MoCa-Qwen25VL-3B](https://huggingface.co/moca-embed/MoCa-Qwen25VL-3B) | [📚 Datasets](https://huggingface.co/moca-embed/datasets) | [📄 Paper](https://arxiv.org/abs/2506.23115)

## Introduction

This is a interleaved multimodal pre-training dataset used in the modality-aware continual pre-training of MoCa models. It is adapted from the train split of [
MMEB](https://huggingface.co/datasets/TIGER-Lab/MMEB-train) by concatenating queries and positive documents.

The dataset consists of interleaved multimodal examples. text is a string containing text while images are image binaries that can be loaded with the following code snippet:

```python
import PIL.Image
from io import BytesIO

image_bytes = example['images'][0]
image = PIL.Image.open(BytesIO(image_bytes))
```


## Citation
MoCa

```bibtex
@article{chen2025moca,
  title={MoCa: Modality-aware Continual Pre-training Makes Better Bidirectional Multimodal Embeddings},
  author={Chen, Haonan and Liu, Hong and Luo, Yuping and Wang, Liang and Yang, Nan and Wei, Furu and Dou, Zhicheng},
  journal={arXiv preprint arXiv:2506.23115},
  year={2025}
}
```

MMEB

```bibtex
@article{jiang2024vlm2vec,
  title={VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks},
  author={Jiang, Ziyan and Meng, Rui and Yang, Xinyi and Yavuz, Semih and Zhou, Yingbo and Chen, Wenhu},
  journal={arXiv preprint arXiv:2410.05160},
  year={2024}
}
```