Datasets:

zhifeixie
/

Voices-in-the-Wild-2M

The default config contains 54 while it should generally contain 3 splits maximum (train/validation/test). If the splits distortion, distortion_dropout, distortion_dropout_noise, distortion_noise, dropout... are not used to differentiate between training and evaluation, please consider defining configs of this dataset instead. You can find how to define configs instead of splits here: https://huggingface.co/docs/hub/datasets-data-files-configuration

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Voices in the Wild

Voices in the Wild is an automatic speech recognition dataset for robustness training and evaluation under diverse acoustic conditions. Audio files use public sequential names and are grouped only by normalized acoustic subset.

Data Fields

file_name: relative path to the audio file.
audio_path: audio path retained for local tooling.
text: transcription alias copied from answer.
answer: reference transcription.
question: transcription instruction.
subset: normalized acoustic condition category.
prediction: empty placeholder for model output.
name: public sample identifier.
index: integer sample index.

Dataset Size

Total examples: 645925
Subset categories: 54

Loading

from datasets import load_dataset, Audio

ds = load_dataset("audiofolder", data_dir="/path/to/Voices-in-the-Wild")
ds = ds.cast_column("audio", Audio())
print(ds["train"][0])

Downloads last month: 3,064