metadata
language:
- bn
size_categories:
- 1M<n<10M
task_categories:
- text-generation
dataset_info:
features:
- name: input_ids
sequence: int32
- name: token_type_ids
sequence: int8
- name: attention_mask
sequence: int8
- name: word_ids
sequence: int64
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 14852870908
num_examples: 5207879
download_size: 3451024663
dataset_size: 14852870908
configs:
- config_name: default
data_files:
- split: train
path: data/train-*