| dataset_info: | |
| features: | |
| - name: idx | |
| dtype: int64 | |
| - name: img_url | |
| dtype: string | |
| - name: image_tokens | |
| sequence: int64 | |
| - name: caption | |
| dtype: string | |
| - name: clean_caption | |
| dtype: string | |
| - name: input_ids | |
| sequence: int32 | |
| - name: attention_mask | |
| sequence: int8 | |
| splits: | |
| - name: train | |
| num_bytes: 944695861 | |
| num_examples: 527187 | |
| download_size: 174001714 | |
| dataset_size: 944695861 | |
| configs: | |
| - config_name: default | |
| data_files: | |
| - split: train | |
| path: data/train-* | |