metadata
dataset_info:
features:
- name: id
dtype: string
- name: dump
dtype: string
- name: url
dtype: string
- name: file_path
dtype: string
- name: language
dtype: string
- name: language_score
dtype: float64
- name: token_count
dtype: int64
- name: score
dtype: float64
- name: int_score
dtype: int64
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 24942837662.25636
num_examples: 5765432
download_size: 13644753644
dataset_size: 24942837662.25636
configs:
- config_name: default
data_files:
- split: train
path: data/train-*