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: 6380218154.8048935
num_examples: 55269252
- name: test
num_bytes: 130208543.19510694
num_examples: 1127944
download_size: 2678498679
dataset_size: 6510426698
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*