metadata
dataset_info:
features:
- name: unit
sequence:
sequence: int64
- name: x-vector
sequence:
sequence: float32
- name: text
dtype: string
- name: text_with_pad
dtype: string
splits:
- name: train
num_bytes: 176657243743
num_examples: 15982064
download_size: 48751951444
dataset_size: 176657243743
configs:
- config_name: default
data_files:
- split: train
path: data/train-*