metadata
dataset_info:
features:
- name: embedding
sequence:
sequence:
sequence:
sequence: float32
- name: label
dtype: int64
splits:
- name: train
num_bytes: 9628715200
num_examples: 20620
download_size: 428533143
dataset_size: 9628715200
configs:
- config_name: default
data_files:
- split: train
path: data/train-*