metadata
dataset_info:
features:
- name: src
dtype: string
- name: gold
dtype: string
- name: rejected
dtype: string
- name: pred
dtype: string
splits:
- name: train
num_bytes: 81488
num_examples: 280
download_size: 54207
dataset_size: 81488
configs:
- config_name: default
data_files:
- split: train
path: data/train-*