metadata
dataset_info:
features:
- name: text
dtype: string
- name: source
dtype: string
- name: emb
sequence: float32
splits:
- name: train
num_bytes: 2354552271
num_examples: 381202
download_size: 2220043562
dataset_size: 2354552271
configs:
- config_name: default
data_files:
- split: train
path: data/train-*