metadata
dataset_info:
features:
- name: input
dtype: string
- name: grund_truth
dtype: string
- name: prediction
dtype: string
- name: mrr
dtype: float64
splits:
- name: train
num_bytes: 214534
num_examples: 500
download_size: 48223
dataset_size: 214534
configs:
- config_name: default
data_files:
- split: train
path: data/train-*