metadata
dataset_info:
features:
- name: index
dtype: int64
- name: query
dtype: string
- name: text
dtype: string
- name: label
dtype: string
- name: preds
dtype: string
- name: average_score
dtype: float64
- name: total_score
dtype: float64
- name: max_score
dtype: float64
- name: min_score
dtype: float64
- name: best_title
dtype: string
- name: clean_preds
dtype: string
- name: new_score
dtype: float64
- name: good_pred
dtype: bool
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 1656074
num_examples: 1629
download_size: 889043
dataset_size: 1656074
configs:
- config_name: default
data_files:
- split: train
path: data/train-*