metadata
dataset_info:
features:
- name: id
dtype: string
- name: query
dtype: string
- name: positives
list:
- name: id
dtype: string
- name: score
dtype: float64
- name: text
dtype: string
- name: negatives
list:
- name: id
dtype: string
- name: score
dtype: float64
- name: text
dtype: string
- name: topk_rank
dtype: int64
splits:
- name: train
num_bytes: 55257110875
num_examples: 631622
download_size: 28690644268
dataset_size: 55257110875
configs:
- config_name: default
data_files:
- split: train
path: data/train-*