metadata
dataset_info:
features:
- name: messages
sequence: string
- name: label
dtype: string
- name: id
dtype: int64
splits:
- name: train
num_bytes: 998000
num_examples: 5000
download_size: 135190
dataset_size: 998000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*