A 1-billion-parameter Norwegian Bokmål speech recognition model fine-tuned from Wav2Vec2 by the National Library of Norway.