metadata
language:
- en
dataset_info:
features:
- name: task
dtype: string
- name: output
dtype: string
- name: instruction
dtype: string
splits:
- name: train
num_bytes: 11136711721.907476
num_examples: 553419
- name: validation
num_bytes: 19878073.715048898
num_examples: 1000
download_size: 5846104069
dataset_size: 11156589795.622526
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*