---
title: moondream2-coyo-5M-captions
canonical_url: "https://www.modelscope.cn/datasets/swift/moondream2-coyo-5M-captions"
md_url: "https://www.modelscope.cn/datasets/swift/moondream2-coyo-5M-captions.md"
repository: swift/moondream2-coyo-5M-captions
last_updated: 2024-08-23
license: cc-by-sa-4.0
storage_size: "1.4 GB"
domain:
  - configs
  - size_categories
  - dataset_info
  - language
  - pretty_name
  - task_categories
tasks:
  - "{config_name=default, data_files=[{split=train, path=data/train-*}]}"
  - "1M<n<10M"
  - dataset_size
  - en
  - "Moondream2 COYO-700M 5M Subset Captions"
  - visual-question-answering
downloads: 3575
stars: 0
---

# moondream2-coyo-5M-captions

> moondream2-coyo-5M-captions - swift 在 ModelScope 开源的数据集。Moondream2 COYO-700M 5M subset captions

swift/moondream2-coyo-5M-captions 是 ModelScope 魔搭社区上的{config_name=default, data_files=[{split=train, path=data/train-*}]}、1M<n<10M、dataset_size数据集，涉及 configs、size_categories、dataset_info 领域，存储大小 1.4 GB，采用 cc-by-sa-4.0 许可。

- **Repository**: swift/moondream2-coyo-5M-captions
- **License**: cc-by-sa-4.0
- **Tasks**: {config_name=default, data_files=[{split=train, path=data/train-*}]}, 1M<n<10M, dataset_size, en, Moondream2 COYO-700M 5M Subset Captions, visual-question-answering
- **Domain**: configs, size_categories, dataset_info, language, pretty_name, task_categories
- **Storage size**: 1.4 GB
- **Downloads**: 3575
- **Stars**: 0
- **Last updated**: 2024-08-23

Source: https://www.modelscope.cn/datasets/swift/moondream2-coyo-5M-captions

---

## Moondream2 COYO-700M 5M subset captions

A 5-million image, text pair subset of COYO-700M dataset, captioned with Moondream2 (rev=`2024-05-08`).  Captioning question is `Write a long caption for this image given the alt text: {alt_text}`.


### Sampling conditions

Randomly sampled from 5 million images from COYO-700M images that fit to the following filters:

```
filters = [
    ("width", ">=", 256),
    ("height", ">=", 256),
    ("aesthetic_score_laion_v2", ">=", 5.2),
    ("watermark_score", "<=", 0.40),
    ("clip_similarity_vitl14", ">=", 0.1),
]
```