Datasets:
metadata
language:
- bn
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 224651981.279
num_examples: 21221
download_size: 291551421
dataset_size: 224651981.279
configs:
- config_name: default
data_files:
- split: train
path: data/train-*