metadata
dataset_info:
features:
- name: prefix
dtype: string
- name: target_id
dtype: int64
- name: label
dtype: string
- name: confidence
dtype: int64
splits:
- name: train
num_bytes: 2441403
num_examples: 20765
download_size: 1156797
dataset_size: 2441403
configs:
- config_name: default
data_files:
- split: train
path: data/train-*