metadata
dataset_info:
features:
- name: output
dtype: string
- name: input
dtype: string
- name: instruction
dtype: string
splits:
- name: train
num_bytes: 11834584.568880351
num_examples: 5101
- name: test
num_bytes: 5073953.431119649
num_examples: 2187
download_size: 7288522
dataset_size: 16908538
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*