|
--- |
|
language: |
|
- vi |
|
- en |
|
pretty_name: "Google Conceptual Captions in Vietnamese" |
|
source-datasets: |
|
- google |
|
tags: |
|
- gcc |
|
- gcc-vi |
|
- Vietnamese |
|
license: unknown |
|
task_categories: |
|
- image-to-text |
|
task_ids: |
|
- image-captioning |
|
--- |
|
# Google Conceptual Captions in Vietnamese |
|
|
|
This is Vietnamese version of Google Conceptual Captions dateset. It has more than 3.3 million image urls with captions. It was built by using Google Translate API. The Vietnamese version has the exact metadata as English one. The only difference is the caption content. |
|
|
|
I provide both English and Vietnamese `.tsv` files. For the English one, one can go to alternative sources: |
|
- https://huggingface.co/datasets/conceptual_captions |
|
- https://github.com/google-research-datasets/conceptual-captions |
|
|
|
To download the dataset, one can use the tool: |
|
- https://github.com/rom1504/img2dataset/blob/main/dataset_examples/cc3m.md |
|
|
|
Or just iterate line by line (`caption<tab>url`) |
|
|
|
⚠ Note: |
|
- Some of image urls might die over the time ([liuhaotian/LLaVA-CC3M-Pretrain-595K](https://huggingface.co/datasets/liuhaotian/LLaVA-CC3M-Pretrain-595K) reported that 15% of original dataset are inaccessible). I'm not responsible for them. |
