metadata
dataset_info:
features:
- name: ds_i
dtype: string
- name: ds_o
dtype: string
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 1010358
num_examples: 100
download_size: 413396
dataset_size: 1010358
configs:
- config_name: default
data_files:
- split: train
path: data/train-*