metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: conv_id
dtype: string
- name: utterance_id
dtype: float32
- name: from
dtype: string
- name: text
dtype: string
- name: emotion
dtype: string
- name: original_full_path
dtype: string
- name: audio_tokens
sequence: int64
splits:
- name: train
num_bytes: 936821457
num_examples: 151644
- name: validation
num_bytes: 51229086
num_examples: 8513
download_size: 154128132
dataset_size: 988050543
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*