metadata
dataset_info:
features:
- name: id
dtype: string
- name: text
dtype: string
- name: embedding
list: float32
- name: url
dtype: string
- name: filename
dtype: string
- name: page_num
dtype: int32
- name: words
dtype: string
- name: corrected_text
dtype: string
- name: corrections
dtype: string
splits:
- name: train
num_bytes: 1572132
num_examples: 22
download_size: 511098
dataset_size: 1572132
configs:
- config_name: default
data_files:
- split: train
path: data/train-*