metadata
dataset_info:
features:
- name: input_ids
sequence:
sequence: int32
- name: attention_mask
sequence:
sequence: int8
- name: labels
sequence:
sequence: int64
splits:
- name: train
num_bytes: 13336000
num_examples: 1000
download_size: 963633
dataset_size: 13336000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*