metadata
dataset_info:
features:
- name: question_text
dtype: string
- name: metadata
struct:
- name: date
dtype: int64
- name: dump
dtype: string
- name: file_path
dtype: string
- name: language
dtype: string
- name: language_score
dtype: float64
- name: token_count
dtype: int64
- name: url
dtype: string
- name: prompt
dtype: string
- name: orig_answer_texts
dtype: int64
splits:
- name: train
num_bytes: 1545615750
num_examples: 200000
download_size: 675180660
dataset_size: 1545615750
configs:
- config_name: default
data_files:
- split: train
path: data/train-*