metadata
license: cc0-1.0
size_categories:
- n<1K
tags:
- embedding
- metadata
pretty_name: Voice Converted Arabic Common Voice 12.0 Speakers
dataset_info:
features:
- name: embedding
sequence: float32
- name: male
dtype: bool
- name: reference
dtype: string
- name: other
sequence: int32
- name: mostly_corrupted
dtype: bool
splits:
- name: train
num_bytes: 27386
num_examples: 26
download_size: 27724
dataset_size: 27386
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
These are the speakers' embeddings and information for the Voice Converted Arabic Common Voice 12.0 dataset.