metadata
dataset_info:
features:
- name: text
dtype: string
- name: embedding
sequence:
sequence: float32
- name: tokens
sequence: int64
- name: database
dtype: string
- name: page
dtype: int64
- name: chunk
dtype: int64
- name: file
dtype: string
splits:
- name: train
num_bytes: 1311360887
num_examples: 144393
download_size: 834129048
dataset_size: 1311360887
configs:
- config_name: default
data_files:
- split: train
path: data/train-*