metadata
dataset_info:
features:
- name: id
dtype: int64
- name: transcript
dtype: string
- name: audio
dtype: audio
- name: intent
dtype: string
- name: 'slots:'
sequence: string
- name: cam
dtype: string
splits:
- name: train
num_bytes: 2922022234.136
num_examples: 50628
- name: test
num_bytes: 709980145.726
num_examples: 13078
- name: devel
num_bytes: 477536022.9
num_examples: 8690
download_size: 3916202128
dataset_size: 4109538402.762
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: devel
path: data/devel-*