metadata
dataset_info:
features:
- name: text
dtype: string
- name: id
dtype: string
- name: probability
dtype: float32
- name: relevant
dtype: bool
- name: dump
dtype: string
- name: url
dtype: string
- name: date
dtype: timestamp[ns]
- name: file_path
dtype: string
- name: language
dtype: string
- name: language_score
dtype: float32
- name: token_count
dtype: int64
- name: idx
dtype: int64
- name: score
dtype: float32
- name: int_score
dtype: int64
- name: source
dtype: string
splits:
- name: train
num_bytes: 12392997275
num_examples: 2644168
download_size: 6804072800
dataset_size: 12392997275
configs:
- config_name: default
data_files:
- split: train
path: data/train-*