Datasets:

avramandrei
/

morovoc

license: mit
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
dataset_info:
  features:
    - name: record_id
      dtype: int32
    - name: dialect
      dtype:
        class_label:
          names:
            '0': moldavian
            '1': romanian
    - name: gender
      dtype:
        class_label:
          names:
            '0': female
            '1': male
    - name: age
      dtype:
        class_label:
          names:
            '0': 10-20
            '1': 20-30
            '2': 30-40
            '3': 40-50
            '4': 50-60
            '5': 50-70
            '6': 60-70
            '7': 70-80
            '8': 80-90
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: sr
      dtype: int32
  splits:
    - name: train
      num_bytes: 18646895232.952
      num_examples: 77638
    - name: validation
      num_bytes: 1421850967.4
      num_examples: 5348
    - name: test
      num_bytes: 1129435874.736
      num_examples: 53485348
  download_size: 21519950558
  dataset_size: 21198182075.088
task_categories:
  - audio-classification
language:
  - ro
tags:
  - dialect
  - political
size_categories:
  - 10K<n<100K

MoRoVoc

MoRoVoc is a dataset that contains audio recordings with the following metadata:

record_id: Unique identifier for each record
dialect: Dialect/language variant of the speaker (i.e., either Romanian or Moldavian)
gender: Gender of the speaker
age: Age range of the speaker
audio: Audio file (WAV format)
sr: Sample rate of the audio

You can read more about the dataset in the following paper: work in progress.

Dataset Structure

Train: 77638 samples
Validation: 5348 samples
Test: 5348 samples

Usage

from datasets import load_dataset

dataset = load_dataset("avramandrei/morovoc")

# Access the splits
train_data = dataset['train']
valid_data = dataset['validation']
test_data = dataset['test']

# Access a sample
sample = train_data[0]
audio_array = sample['audio']['array']
sampling_rate = sample['audio']['sampling_rate']

Citation

Work in progress