---
dataset_info:
  features:
  - name: observation.images.image
    dtype: video_frame
  - name: language_instruction
    dtype: string
  - name: observation.state
    sequence: float32
    length: 12
  - name: action
    sequence: float32
    length: 7
  - name: timestamp
    dtype: float32
  - name: episode_index
    dtype: int64
  - name: frame_index
    dtype: int64
  - name: next.reward
    dtype: float32
  - name: next.done
    dtype: bool
  - name: index
    dtype: int64
  splits:
  - name: train
    num_bytes: 1708281
    num_examples: 7622
  download_size: 888370
  dataset_size: 1708281
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---