Self-supervised approach for learning speech representations in a sustainable way - View it on GitHub