I wonder if you could generate synthetic Victorian-era training data.
Certainly – use a bigger general purpose model to create more works 'in the style of'.
Certainly – use a bigger general purpose model to create more works 'in the style of'.