metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: id
dtype: string
- name: text
dtype: string
- name: topic
dtype: string
- name: field
dtype: string
- name: subfield
dtype: string
- name: domain
dtype: string
- name: abstract
dtype: string
- name: lang
dtype: string
- name: sentences
list: string
- name: embeddings
list:
list: float32
- name: sem_embeddings
list:
list: float32
- name: matches
list:
list:
- name: m
dtype: string
- name: s
dtype: string
splits:
- name: train
num_bytes: 40588484561
num_examples: 779932
download_size: 39735445995
dataset_size: 40588484561