metadata
dataset_info:
features:
- name: id
dtype: string
- name: title
dtype: string
- name: text
dtype: string
- name: embeddings
sequence: float32
splits:
- name: train
num_bytes: 117390606165
num_examples: 19399177
download_size: 107683349989
dataset_size: 117390606165
configs:
- config_name: default
data_files:
- split: train
path: data/train-*