metadata
dataset_info:
features:
- name: episode_id
dtype: string
- name: step_id
dtype: int64
- name: screenshot_image
dtype: image
- name: source
dtype: string
- name: os
dtype: string
- name: goal
dtype: string
- name: action_history
dtype: string
- name: current_question
dtype: string
- name: options
list: string
- name: answer
dtype: int64
splits:
- name: train
num_bytes: 366714274.9
num_examples: 1698
download_size: 343047168
dataset_size: 366714274.9
configs:
- config_name: default
data_files:
- split: train
path: data/train-*