> We don't have much in the way of recorded speech from that area We may not have a ton, but do have a lot of news reels and radio broadcasts from the time sur

> We don't have much in the way of recorded speech from that area We may not have a ton, but do have a lot of news reels and radio broadcasts from the time surrounding WWI. Certainly enough to style-transfer a voice model to plug into the text model.