utokyo_saytap_converted_externally_to_rlds

FeaturesDict({
    'episode_metadata': FeaturesDict({
        'file_path': Text(shape=(), dtype=string),
    }),
    'steps': Dataset({
        'action': Tensor(shape=(12,), dtype=float32, description=Robot action, consists of [12x joint positios].),
        'discount': Scalar(shape=(), dtype=float32, description=Discount if provided, default to 1.),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'language_embedding': Tensor(shape=(512,), dtype=float32, description=Kona language embedding. See https://tfhub.dev/google/universal-sentence-encoder-large/5),
        'language_instruction': Text(shape=(), dtype=string),
        'observation': FeaturesDict({
            'desired_pattern': Tensor(shape=(4, 5), dtype=bool, description=Desired foot contact pattern for the 4 legs, the 4 rows are for the front right, front left, rear right and rear left legs, the pattern length is 5 (=0.1s).),
            'desired_vel': Tensor(shape=(3,), dtype=float32, description=Desired velocites. The first 2 are linear velocities along and perpendicular to the heading direction, the 3rd is the desired angular velocity about the yaw axis.),
            'image': Image(shape=(64, 64, 3), dtype=uint8, description=Dummy camera RGB observation.),
            'prev_act': Tensor(shape=(12,), dtype=float32, description=Actions applied in the previous step.),
            'proj_grav_vec': Tensor(shape=(3,), dtype=float32, description=The gravity vector [0, 0, -1] in the robot base frame.),
            'state': Tensor(shape=(30,), dtype=float32, description=Robot state, consists of [3x robot base linear velocity, 3x base angular vel, 12x joint position, 12x joint velocity].),
            'wrist_image': Image(shape=(64, 64, 3), dtype=uint8, description=Dummy wrist camera RGB observation.),
        }),
        'reward': Scalar(shape=(), dtype=float32, description=Reward if provided, 1 on final step for demos.),
    }),
})

@article{saytap2023,
  author = {Yujin Tang and Wenhao Yu and Jie Tan and Heiga Zen and Aleksandra Faust and
Tatsuya Harada},
  title  = {SayTap: Language to Quadrupedal Locomotion},
  eprint = {arXiv:2306.07580},
  url    = {https://saytap.github.io},
  note   = "{https://saytap.github.io}",
  year   = {2023}
}