metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: context
dtype: string
- name: response
dtype: string
- name: category
dtype: string
- name: text
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 19714838
num_examples: 12008
download_size: 12073239
dataset_size: 19714838
configs:
- config_name: default
data_files:
- split: train
path: data/train-*