Modelpublic

Africanvoice/african_voices_turbo_hausa

sourceHugging Faceapache-2.0updated 5d agoView on Hugging Face

0likes23downloads

Model Card

Whisper Turbo finetuned using African Voices Hausa Dataset

This model is a fine-tuned version of openai/whisper-large-v3-turbo for Hausa Language.

Model Description

Whisper Turbo finetuned using African Voices Hausa Dataset is an automatic speech recognition (ASR) model trained on Africanvoice/African_voices_hausa dataset.

Performance

—eval_loss: 0.26
—WER: 12.81%

Training Details

—Epochs: 9
—Batch Size: 24
—Learning Rate: 1e-6

Usage

python

from transformers import pipeline

# Load the model
pipe = pipeline("automatic-speech-recognition", model="Africanvoice/african_voices_turbo_hausa")

# Transcribe audio
result = pipe("path/to/audio.wav")
print(result["text"])

Citation

If you use this model, please cite:

bibtex

@misc{Whisper Turbo finetuned using African Voices Hausa Dataset,
  author = {EqualyzAI and African Voices Team},
  title = {Whisper Turbo finetuned using African Voices Hausa Dataset},
  year = {2025},
  publisher = {HuggingFace},
  howpublished = {\url{https://huggingface.co/Africanvoice/african_voices_turbo_hausa}}
}