metadata
dataset_info:
features:
- name: '0'
dtype: string
- name: embeddings
sequence: float32
splits:
- name: train
num_bytes: 35697138
num_examples: 19021
download_size: 44362425
dataset_size: 35697138
configs:
- config_name: default
data_files:
- split: train
path: data/train-*