metadata
dataset_info:
features:
- name: contexts
list: string
- name: question
dtype: string
- name: answer
dtype: string
- name: ctx_len
dtype: int64
- name: context_token_count
dtype: int64
splits:
- name: train
num_bytes: 806183371
num_examples: 59886
download_size: 332187434
dataset_size: 806183371
configs:
- config_name: default
data_files:
- split: train
path: data/train-*