Dataset + policies + eval for the **long_horizon1** task (WetLabRoboData / ChemLab imitation-learning benchmark).