Judging from the chart in their github README, Whisper performs much better in parsing Spanish audio than any other language and that in particular blows my min

Judging from the chart in their github README, Whisper performs much better in parsing Spanish audio than any other language and that in particular blows my mind. I would have expected English to be at the top of any such model, it being such an IT lingua franca. Now I wonder if it works equally well with Spanish from Spain (and its different regions) and Spanish from the New World (and in its myriads of different flavours).