metadata
dataset_info:
features:
- name: name
dtype: string
- name: embd
sequence: float32
splits:
- name: train
num_bytes: 20338743
num_examples: 9810
download_size: 26269064
dataset_size: 20338743
configs:
- config_name: default
data_files:
- split: train
path: data/train-*