metadata
dataset_info:
features:
- name: dataset
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: prompt_len
dtype: int64
splits:
- name: train
num_bytes: 137340482
num_examples: 44694
download_size: 76395752
dataset_size: 137340482
configs:
- config_name: default
data_files:
- split: train
path: data/train-*