---
dataset_info:
  features:
  - name: episode_index
    dtype: int64
  - name: actions
    dtype: int64
  - name: text_actions
    dtype: string
  - name: plans
    dtype: string
  - name: rewards
    dtype: float64
  - name: terms
    dtype: bool
  - name: truncs
    dtype: bool
  - name: short_term_context
    dtype: string
  - name: long_term_context
    dtype: string
  - name: image_paths
    dtype: image
  splits:
  - name: train
    num_bytes: 404681541.814
    num_examples: 18183
  download_size: 389881532
  dataset_size: 404681541.814
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---