Closed lkiesow closed 1 year ago
Fixed merge conflict
Updated this.
You can now specify multiple models (-m model1 model2
) or languages (-l lang1 lang2
) or use the special value auto
. The latter should be especially handy for the Opencast integration since even with the current service/woh you should already be able to just specify the language as auto
.
This patch allows users to simply let
vosk-cli
auto-detect the best model to use for the provided audio file by probing the first part of the audio against the available models.