metadata
license: cc-by-3.0
dataset_info:
features:
- name: output
dtype: string
- name: qid
dtype: string
- name: name
dtype: string
- name: input
dtype: string
- name: instruction
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 1854503569
num_examples: 1560888
download_size: 1048410730
dataset_size: 1854503569
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
language:
- en
tags:
- wikidata
- wikipedia
pretty_name: Wikidata English descriptions dataset (downsmapled)
size_categories:
- 1M<n<10M