metadata
dataset_info:
features:
- name: observation.point_cloud
dtype:
array2_d:
shape:
- 1000
- 3
dtype: float32
- name: observation.state
dtype:
array2_d:
shape:
- 1
- 3
dtype: float32
- name: action
dtype:
array2_d:
shape:
- 1
- 3
dtype: float32
- name: episode_index
dtype: int64
- name: next.done
dtype: bool
splits:
- name: train
num_bytes: 16437376
num_examples: 1024
download_size: 12051469
dataset_size: 16437376
configs:
- config_name: default
data_files:
- split: train
path: data/train-*