metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: instruction
dtype: string
- name: context
dtype: string
- name: response
dtype: string
- name: category
dtype:
class_label:
names:
'0': general_qa
'1': closed_qa
'2': open_qa
'3': classification
'4': information_extraction
splits:
- name: train
num_bytes: 99456
num_examples: 64
download_size: 12865
dataset_size: 99456