metadata
dataset_info:
features:
- name: input_length
dtype: int64
- name: input_features
sequence:
sequence: float32
- name: labels
sequence: int64
- name: labels_length
dtype: int64
splits:
- name: train
num_bytes: 17561799568
num_examples: 18219
- name: valid
num_bytes: 1951953408
num_examples: 2025
download_size: 19403201935
dataset_size: 19513752976
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*