evageon/myaudio

Please note this dataset is private

Using the data

You can stream the data data loader:

myaudio = load_dataset(
    "evageon/myaudio",
    use_auth_token=os.environ["HG_USER_TOKEN"], # replace this with your access token
    streaming=True)

Then you can iterate over the dataset

# replace test with validation or train depending on split you need
print(next(iter(myaudio["test"])))

outputs:

{'path': 'CD93A8FF-C3ED-4AD4-95A6-8363CCB93B90_spk-0001_seg-0024467:0025150.wav', 'audio': {'path': 'dataset/test/wav/CD93A8FF-C3ED-4AD4-95A6-8363CCB93B90_spk-0001_seg-0024467:0025150.wav', 'array': array([0.00662231, 0.00497437, 0.00518799, ..., 0.01150513, 0.00708008,
       0.00296021]), 'sampling_rate': 16000}, 'text': 'خطرا على دول الخليج لماذا اعتبرت أن إيران اليوم والخطر الذي تشكله إيران مختلف'}