metadata
dataset_info:
features:
- name: raw_content
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 6763101375
num_examples: 32000
download_size: 2956465112
dataset_size: 6763101375