![]() There are also whisper videos on youtube and the following link talks about using whisper. ![]() ![]() You can also consider skipping the user interface of that project and use one of the transcribing projects that it uses directly ( the default whisper model can take twice as long as the audio file from what I saw online and whisper.cpp seems to be the most popular modification at the time of writing and seems fairly fast). I found a Github software project that is powered by the different modifications of whisper but I did not try to use it. It also has a built in editor which is pretty handy for quick editing like removing filler words or random sounds like cheering etc. I am currently searching for this and so far it seems that openai's whisper (note that there seems to be both an api call and an option for a local installation) and its modifications are the best solutions as they seem to be both open source and fairly accurate (though not 100%). Its free and the transcription quality is the best I have ever found so far. The free, offline, fast, faithful way is to use Python, there are many tutorials that explain how to do it, both for STT and for TTS, for the audio files you have, you can virtualize the microphone with a suitable program such as vb-audio cable.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |