metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: text
dtype: string
- name: score
dtype: float32
- name: expert_name
dtype: string
- name: label
dtype: string
splits:
- name: train
num_bytes: 13590789445.868557
num_examples: 125177457
- name: test
num_bytes: 277363052.1314436
num_examples: 2554642
download_size: 6697715035
dataset_size: 13868152498
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*