metadata
dataset_info:
features:
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 217102402
num_examples: 17642
download_size: 99771216
dataset_size: 217102402
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
A mix of part of datasets: pubmed, pubmed_qa and alpaca