| dataset_info: | |
| features: | |
| - name: id | |
| dtype: string | |
| - name: type | |
| dtype: string | |
| - name: video | |
| dtype: string | |
| - name: input | |
| dtype: string | |
| - name: length | |
| dtype: int64 | |
| - name: answer | |
| dtype: string | |
| splits: | |
| - name: test | |
| num_bytes: 634664 | |
| num_examples: 1500 | |
| download_size: 62857 | |
| dataset_size: 634664 | |
| configs: | |
| - config_name: default | |
| data_files: | |
| - split: test | |
| path: data/test-* | |
| If you use TimeScope please cite the following: | |
| ```bibtex | |
| @misc{zohar2025apollo2, | |
| title = {Apollo2: Exploring the Long-Video Frontier of Large Multimodal Models}, | |
| author = {Zohar, Orr and Wang, Xiaohan and Li, Rui and Marafioti, Andrés and Farré, Miquel and Noyan, Merve and von Werra, Leandro and Yeung-Levy, Serena and Wolf, Thomas}, | |
| year = {2025}, | |
| } | |
| ``` | |