---
dataset_info:
  features:
  - name: task_id
    dtype: int64
  - name: seq_data
    struct:
    - name: actions
      list:
        list: float32
    - name: done
      list: bool
    - name: features
      struct:
      - name: out_features
        list:
          list: float32
    - name: obs
      struct:
      - name: state
        list:
          list: float32
  splits:
  - name: train
    num_bytes: 7473484500
    num_examples: 500
  download_size: 6522986972
  dataset_size: 7473484500
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---