| license: mit | |
| task_categories: | |
| - robotics | |
| tags: | |
| - robotics | |
| - robomme | |
| - manipulation | |
| - memory | |
| - imitation-learning | |
| - hdf5 | |
| size_categories: | |
| - n<1K | |
| # RoboMME — SequenceRecoveryVertically (Robot HDF5 Demonstrations) | |
| Raw robot demonstration data for the **SequenceRecoveryVertically** task from | |
| [RoboMME](https://github.com/), a ManiSkill/SAPIEN benchmark for | |
| memory-augmented robotic manipulation. Each HDF5 file is one recorded episode | |
| containing observations (RGB/state), actions, and metadata for imitation | |
| learning. | |
| ## Layout | |
| ``` | |
| train/ 100 episodes | |
| val/ 50 episodes | |
| test/ 50 episodes | |
| ``` | |
| Files are named `episode_<idx>_seed_<seed>.h5`. Seeds are baked into the | |
| RoboMME metadata for reproducibility. See `doc/h5_data_format.md` in the RoboMME | |
| repository for the full dataset schema. | |