Closed JanChec closed 1 year ago
Hey, this seems like something that could be handled by a user configuration which replaces or removes NSFW words.
This should be fairly easy to do, check the current examples, it only requires word replacement which is quit simple.
Closing as this can be handled as part of user configuration, further it would mean having a list of NSFW words which is unlikely to be generally agreed on as it differs between English speaking countries and also depend on context. So I rather let users handle their own word-replacement tables.
NOTE: it would be better to handle this at a language model level as it means NSFW words wont be accidentally selected and other words will be prioritized (instead of NSFW words being ignored or substituted). Submitted: https://github.com/alphacep/vosk-api/issues/1340
Hey! :slightly_smiling_face: I'm using your amazing QoL improvement for work. Sometimes if it mishears me it produces some other word. It happened to me to include a scrambled "fcuk". Fortunately not during a meeting :laughing: I'd love to be able to use it during meetings to speed up any writing down that is typically a bottleneck, but I'd like to not end up with NSFW situation :smile: