metadata
dataset_info:
features:
- name: text
dtype: string
- name: embd
sequence: float64
- name: cluster
sequence: float64
- name: doc_ids
dtype: string
- name: level_id
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 6769749
num_examples: 882
download_size: 6037364
dataset_size: 6769749
configs:
- config_name: default
data_files:
- split: train
path: data/train-*