Sherlockhu/drone-nav-synthetic-v1

该数据集包含图像、动作、episode_id、step_idx、embodiment、is_first、is_last和language_instruction等特征。数据集只有一个名为train的分割，包含2500个样本，总大小为239368013字节。该数据集可能用于涉及图像和动作的任务，如机器人或AI训练，特别是考虑到action和language_instruction特征的存在。The dataset includes features such as image, action, episode_id, step_idx, embodiment, is_first, is_last, and language_instruction. It has a single split named train with 2500 examples and a total size of 239368013 bytes. The dataset is likely intended for tasks involving images and actions, such as robotic or AI training, given the presence of action and language_instruction features.