metadata
language:
- en
dataset_info:
features:
- name: input
dtype: string
- name: history
sequence: 'null'
- name: output
sequence: string
- name: instruction
dtype: string
splits:
- name: train
num_bytes: 1334567079
num_examples: 615602
download_size: 129152712
dataset_size: 1334567079
configs:
- config_name: default
data_files:
- split: train
path: data/train-*