metadata
dataset_info:
features:
- name: input
sequence: string
- name: output
sequence: string
- name: instructions
dtype: string
splits:
- name: train
num_bytes: 3264597
num_examples: 2685
- name: validation
num_bytes: 468311
num_examples: 383
- name: test
num_bytes: 941957
num_examples: 768
download_size: 400161
dataset_size: 4674865
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*