metadata
dataset_info:
features:
- name: conversations
struct:
- name: from
dtype: string
- name: value
dtype: string
- name: category
dtype: string
splits:
- name: train
num_bytes: 33014035
num_examples: 63989
download_size: 13421243
dataset_size: 33014035
configs:
- config_name: default
data_files:
- split: train
path: data/train-*