| dataset_info: | |
| features: | |
| - name: images | |
| sequence: image | |
| - name: texts | |
| list: | |
| - name: user | |
| dtype: string | |
| - name: assistant | |
| dtype: string | |
| - name: source | |
| dtype: string | |
| splits: | |
| - name: train | |
| num_bytes: 244951255658.16818 | |
| num_examples: 565009 | |
| download_size: 145422811605 | |
| dataset_size: 244951255658.16818 | |
| configs: | |
| - config_name: default | |
| data_files: | |
| - split: train | |
| path: data/train-* | |
| [Docmatix](https://huggingface.co/datasets/HuggingFaceM4/Docmatix), but with multi-image samples filtered out. |