Support for Italian language?
#12
by ManuSri - opened
Hello, I was using this model last month for transcribing Italian speech and it was working fine.
Today I was checking the hugging face repo and it says that it supports just German and English and I am bit confused.
Has it always been only for two languages?
Since it was only finetuned on English and German Data performance and verbatimness is likely better in these languages. If you however observe that it works well in italian too that makes me very happy. It would mean that these capabilities and the retokenization seem to (atleast partly) transfer to other languages too even tough the model has not specifically been trained on languages other than english and german after the retokenization.
thank you for the clarification!