metadata
dataset_info:
features:
- name: chunk_id
dtype: string
- name: text
dtype: string
- name: metadata
struct:
- name: Header 1
dtype: string
- name: Header 2
dtype: string
- name: image_references
sequence: string
- name: images_base64
sequence: string
- name: start_index
dtype: int64
- name: source_filename
dtype: string
splits:
- name: train
num_bytes: 1678830
num_examples: 43
download_size: 1196302
dataset_size: 1678830
configs:
- config_name: default
data_files:
- split: train
path: data/train-*