metadata
dataset_info:
features:
- name: id
dtype: string
- name: path
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: duration
dtype: float32
- name: language
dtype: string
- name: original_speaker_id
dtype: int64
- name: session_id
dtype: int64
- name: topic
dtype: string
splits:
- name: train
num_bytes: 239635153.8996356
num_examples: 2331
- name: validation
num_bytes: 24756134.12920354
num_examples: 262
- name: test
num_bytes: 20749586.90570342
num_examples: 257
download_size: 230277242
dataset_size: 285140874.93454254
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*