metadata
dataset_info:
features:
- name: inputs
dtype: string
- name: targets
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: annotation_type
dtype: string
- name: user_id
dtype: string
splits:
- name: train
num_bytes: 1990086
num_examples: 4909
download_size: 981588
dataset_size: 1990086
configs:
- config_name: default
data_files:
- split: train
path: data/train-*