metadata
dataset_info:
features:
- name: user_id
dtype: string
- name: path
dtype: string
- name: text
dtype: string
- name: intent
dtype: string
- name: audio
struct:
- name: array
sequence: float64
- name: path
dtype: string
- name: sampling_rate
dtype: int64
- name: sentence
dtype: string
- name: input_features
sequence:
sequence: float32
- name: token_ids
sequence: int64
splits:
- name: train
num_bytes: 43905796552
num_examples: 32876
- name: test
num_bytes: 18822791942
num_examples: 14090
download_size: 10402644535
dataset_size: 62728588494
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*