---
dataset_info:
  features:
  - name: metadata
    dtype: string
  - name: info
    dtype: string
  - name: observation
    dtype: image
  - name: action
    dtype: int64
  - name: reward
    dtype: float32
  - name: done
    dtype: bool
  - name: episode_id
    dtype: int64
  - name: step_id
    dtype: int64
  splits:
  - name: train
    num_bytes: 381343
    num_examples: 300
  download_size: 41486
  dataset_size: 381343
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---