metadata
dataset_info:
features:
- name: text
dtype: string
- name: response
dtype: string
- name: instruction
dtype: string
splits:
- name: train
num_bytes: 1718148855
num_examples: 1464006
download_size: 746958255
dataset_size: 1718148855
configs:
- config_name: default
data_files:
- split: train
path: data/train-*