metadata
dataset_info:
features:
- name: text
dtype: string
- name: id
dtype: string
- name: dump
dtype: string
- name: url
dtype: string
- name: date
dtype: string
- name: file_path
dtype: string
- name: language
dtype: string
- name: language_score
dtype: float64
- name: language_script
dtype: string
- name: minhash_cluster_size
dtype: int64
- name: top_langs
dtype: string
splits:
- name: train
num_bytes: 21009399391
num_examples: 3000000
download_size: 10370943887
dataset_size: 21009399391
configs:
- config_name: default
data_files:
- split: train
path: data/train-*