it seems that there have always been 3 of them, and for what reason it is not clear, but they are almost the same.small differences
They are different Whisper models: large-v1, large-v2, and large-v3
The subs always seem to be 10-15 seconds behind, am I doing something wrong or is it just the limitations of the technology?
is what i use it's free, just need to normalize then extract the audio, then upload the audio on google drive then do the process on google collab
3 hours video like canan's take about 45~60 minutes to finish and the accuracy is highly dependent on audio quality and accent/pronunciation of the speaker
Last edited:
Which transmission are you referring to? In which one he shouts like "stupid" (I just had doubts)
It's really funny how youtube Once recommended me a Canan stream then once it ended I was immediately transferred to an archived Noel ASMR stream. Youtube knows what's up lmao
Rather , the trends are just set up that way , but it 's a good joke.
