metadata
dataset_info:
features:
- name: tags
dtype: string
- name: instruction
dtype: string
- name: category
dtype: string
- name: intent
dtype: string
- name: response
dtype: string
- name: pure_unicode
dtype: string
- name: mixed_unicode
dtype: string
- name: romanized
dtype: string
- name: pure_unicode_rev
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 4041930
num_examples: 2775
download_size: 1005246
dataset_size: 4041930
configs:
- config_name: default
data_files:
- split: train
path: data/train-*