Datasets:
Modalities:
Text
Formats:
parquet
Languages:
French
Size:
10M - 100M
Tags:
administrative documents
whole word masking
text-mining
text generation
information extraction
License:
metadata
license: cc-by-nc-sa-4.0
dataset_info:
features:
- name: Sentences
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 3667301553
num_examples: 10065432
download_size: 2239239343
dataset_size: 3667301553
configs:
- config_name: default
data_files:
- split: train
path: data/train-*