metadata
dataset_info:
features:
- name: env
dtype: string
- name: prompt
dtype: string
- name: extra
struct:
- name: game_data
struct:
- name: answer
dtype: string
- name: difficulty
dtype: int64
- name: gpt_response
dtype: string
- name: metadata
struct:
- name: depth
dtype: int64
- name: language
dtype: string
- name: num_options
dtype: int64
- name: seed
dtype: int64
- name: true_count
dtype: int64
- name: question
dtype: string
- name: metadata
struct:
- name: depth
dtype: int64
- name: language
dtype: string
- name: num_options
dtype: int64
- name: seed
dtype: int64
- name: true_count
dtype: int64
- name: seed
dtype: int64
- name: task_id
dtype: int64
- name: task_type
dtype: string
splits:
- name: train
num_bytes: 12500120
num_examples: 4019
download_size: 9024955
dataset_size: 12500120
configs:
- config_name: default
data_files:
- split: train
path: data/train-*