metadata
dataset_info:
features:
- name: doc_content
dtype: string
- name: fetch_time
dtype: string
- name: url
dtype: string
- name: warc_filename
dtype: string
- name: warc_record_length
dtype: float64
- name: warc_record_offset
dtype: float64
splits:
- name: train
num_bytes: 10625934878.833914
num_examples: 1424409
download_size: 3185855365
dataset_size: 10625934878.833914
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
language:
- ka