metadata
dataset_info:
features:
- name: output
dtype: string
- name: input
dtype: string
- name: our_inference
dtype: string
- name: our_inference_after_alignment
dtype: string
splits:
- name: train
num_bytes: 787924
num_examples: 742
download_size: 377323
dataset_size: 787924
configs:
- config_name: default
data_files:
- split: train
path: data/train-*