metadata
configs:
- config_name: default
data_files:
- split: test
path: data/test-*
- split: train
path: data/train-*
- split: validation
path: data/validation-*
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype:
class_label:
names:
'0': financial_reports
'1': government_tenders
'2': manuals
'3': laws_and_regulations
'4': scientific_articles
'5': patents
- name: ground_truth
dtype: string
splits:
- name: test
num_bytes: 3240643
num_examples: 12
- name: train
num_bytes: 16492390
num_examples: 43
- name: validation
num_bytes: 1929905.3125
num_examples: 5
download_size: 21721061
dataset_size: 21662938.3125