metadata
dataset_info:
features:
- name: input_ids
list: int32
length: 5120
- name: token_count
dtype: int32
- name: source_file
dtype: string
- name: split
dtype: string
splits:
- name: train
num_bytes: 36914468
num_examples: 1786
- name: validation
num_bytes: 10375721
num_examples: 502
download_size: 46921559
dataset_size: 47290189
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*