metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: caption
dtype: string
- name: emotions
dtype: string
- name: raw_json
dtype: string
splits:
- name: train
num_bytes: 87333539710.584
num_examples: 2179738
download_size: 84125498845
dataset_size: 87333539710.584
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Still a WIP, full dataset is still being annotated