metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: wav
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: emotion
dtype:
class_label:
names:
'0': happy
'1': sad
'2': angry
'3': other
- name: duration
dtype: float32
- name: emotion_start
dtype: float32
- name: emotion_end
dtype: float32
- name: speaker_id
dtype: string
splits:
- name: train
num_bytes: 30997981
num_examples: 180
download_size: 30999146
dataset_size: 30997981
configs:
- config_name: default
data_files:
- split: train
path: data/train-*