Amharic speech to text transcription

Data

This repo uses public dataset provided by ALFFA_PUBLIC https://github.com/besacier/ALFFA_PUBLIC/tree/master/ASR/AMHARIC

Major steps in Speech to text ranscription

• Audio preprocessing
• Audio and text feature extraction
• Modeling and
• Evaluation

In this repository we have implemented all thoe steps

Our repository structure

• .dvc folder contains our versioned data information
• scripts folder contains all python file we have you used to implement amharic speech to text package
• test folder contains test implemntation for some of the functions