---
dataset_info:
  features:
  - name: audio
    dtype: audio
  - name: ipa
    dtype: string
  - name: text
    dtype: string
  - name: g2p
    dtype: string
  - name: speaker_code
    dtype: string
  splits:
  - name: train
    num_bytes: 189253400
    num_examples: 1894
  - name: test
    num_bytes: 127689987
    num_examples: 1266
  download_size: 276061676
  dataset_size: 316943387
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
license: cc-by-nc-4.0
task_categories:
- automatic-speech-recognition
language:
- en
tags:
- Argentina
- Speech
- IPA
pretty_name: EpaDB
size_categories:
- 1K<n<10K
---

# EpaDB
EpaDB is a speech database of 50 native Spanish speakers (25 male, 25 female) from Argentina speaking English.
It contains phonemic annotations using mainly the sounds supported by [ARPABet](https://en.wikipedia.org/wiki/ARPABET) with a few extensions to model Spanish influenced dialects of English.
It was developed by Jazmin Vidal, Luciana Ferrer, and Leonardo Brambilla at the [Speech Lab](http://habla.dc.uba.ar/). Read more on their [official github](https://github.com/JazminVidal/EpaDB) and [paper](https://github.com/JazminVidal/EpaDB/blob/master/EpaDB%20-%20Paper.pdf).

## This Processed Version
We have processed the dataset into an easily consumable [Hugging Face dataset](https://huggingface.co/docs/datasets/en/index) using [this data processing script](https://github.com/KoelLabs/ML/blob/main/scripts/data_loaders/EpaDB.py).
This maps the EpaDB phoneme annotations to [IPA](https://en.wikipedia.org/wiki/International_Phonetic_Alphabet) as supported by libraries like [ipapy](https://pypi.org/project/ipapy/0.0.1.0/) and [panphon](https://pypi.org/project/panphon/0.5/).
We also correct the occasional typos or formatting inconsistencies.

- The train set has 1894 samples (around 98 minutes of speech).
- The test set has 1266 samples (around 66 minutes of speech).

All audio has been converted to float32 in the -1 to 1 range at 16 kHz sampling rate.

## Usage
0. Request access to [this dataset](https://huggingface.co/datasets/KoelLabs/EpaDB) on the Hugging Face website. You will be automatically approved upon accepting the terms.
1. `pip install datasets`
2. [Login to Hugging Face](https://huggingface.co/docs/huggingface_hub/en/guides/cli#huggingface-cli-login) using `huggingface-cli login` with a token that has gated read access.
3. Use the dataset in your scripts:
```python
from datasets import load_dataset

EpaDB = load_dataset("KoelLabs/EpaDB")
train_ds = EpaDB['train']
test_ds = EpaDB['test']

sample = train_ds[0]
print(sample)
```

## License
The original dataset is released under the CC BY-NC 4.0 license, a summary of the license can be found [here](https://creativecommons.org/licenses/by-nc/4.0/), and the full license can be found [here](https://creativecommons.org/licenses/by-nc/4.0/legalcode). 
This processed dataset follows the same license.

For any usage that is not covered by the CC BY-NC 4.0 license, you must contact the [original dataset authors](https://github.com/JazminVidal/EpaDB?tab=readme-ov-file#license).