metadata
dataset_info:
features:
- name: chunk
dtype: string
- name: source
dtype: string
- name: embeddings
sequence: float32
splits:
- name: train
num_bytes: 11364916
num_examples: 4659
download_size: 11216643
dataset_size: 11364916
configs:
- config_name: default
data_files:
- split: train
path: data/train-*