metadata
dataset_info:
features:
- name: content
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 206266358721
num_examples: 21702269
download_size: 51202037064
dataset_size: 206266358721
configs:
- config_name: default
data_files:
- split: train
path: data/c-sharp/train-*