metadata
dataset_info:
features:
- name: text
dtype: string
- name: output
dtype: string
- name: instruction
dtype: string
splits:
- name: train
num_bytes: 42743266
num_examples: 8883
- name: valid
num_bytes: 10729544
num_examples: 2221
- name: test
num_bytes: 6828912
num_examples: 1490
download_size: 23255621
dataset_size: 60301722
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*
- split: test
path: data/test-*