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
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 66307133402
num_examples: 9490000
- name: validation
num_bytes: 1271646268
num_examples: 182101
download_size: 17514445942
dataset_size: 67578779670
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*