A 1 billion parameter Wav2Vec2 model fine-tuned for automatic speech recognition in Norwegian Nynorsk.