his repository contains an Automatic Speech Recognition (ASR) system for Amharic built by fine-tuning Facebook’s Wav2Vec2.0 model using Hugging Face Transformers. The goal is to provide an open-source Amharic speech-to-text model, making it easier for developers and researchers to work with Amharic audio data.
Amharic Speech-to-Text (ASR) with Wav2Vec2
This repository provides an Automatic Speech Recognition (ASR) system for Amharic built using Facebook’s Wav2Vec2.0 model fine-tuned on Amharic speech datasets. The project aims to make Amharic speech recognition accessible and open-source, supporting low-resource language research and real-world applications.
Clone the repo and install dependencies:
git clone
github.com
pip install
Main dependencies:
Python 3.9+
PyTorch
Transformers
Datasets
Accelerate
jiwer (for WER)
📂 Dataset
This project uses Amharic speech datasets. You can replace or expand with your own data.
Example structure:
dataset/
├── train/
│ ├── audio1.wav
│ ├── audio2.wav
│ └── ...
├── test/
│ ├── audio1.wav
│ └── ...
└── metadata.csv # transcripts, speaker info, etc.