Can you share how you run the whisper app, do you run it locally or on virtual cloud? I tried using that for some youtube videos but it takes forever even with large model to translate 10 min video.
Can you share how you run the whisper app, do you run it locally or on virtual cloud? I tried using that for some youtube videos but it takes forever even with large model to translate 10 min video.
Hello, I'm using the cloud version, I really don't even know how I got it myself, because first I had to install 1 python 3.10 2. git 3. Ffmeg on my computer, since according to the video I saw it was still necessary using the cloud version, after installing, so to speak, the drivers requested by the creator of the Collab (Whisper+Vad) it is necessary to install each time we use this Collab, I think on page 219 there is the address of the Collab that comes improved with VAD, too leave some images with indications of what I did before using the Collab.Can you share how you run the whisper app, do you run it locally or on virtual cloud? I tried using that for some youtube videos but it takes forever even with large model to translate 10 min video.
Hi,In my experience, using the virtual version of whisper, with the use of the large model. It takes about 5mins to 10mins to transcribe a 60mins long video. I don't have the powerful GPU installed on my computer, which is why I am also using the virtual version. Using the virtual version, you don't need to install anything on your computer. I did not install anything from my computer because I don't know how to install it honestly. In the virtual version, all you need to do is to upload, select the language, select the model then wait for the transcription to finish.
Appreciate your contribution and welcome hereHello, guys! I'm new to the forum and I already would like to make my contribution
I translated RCTD-111 - Super Gachinko Lesbian Battle DYNAMITE 2018 using lots of different tecniques which include machine translation and my own knowledge of japanese (I'm learning). I think the results were 85% accurate with pretty little mistakes, I got a little sloppy at the end, but understand that this is a 3h movie so it's understandable that I got a bit tired after this goliath of a project.
I used the 720p [3:14:44] version of the movie, please feel free to correct any mistakes that you find.
I will be posting here a lot, mainly subtitles for lesbian movies because they don't get much love from translators, but I will post a few subs for hetero movies I like as well. The next movies on my schedule are:
- BBAN-086
- MIAE-208
- BBAN-242
- LZDM-002
Bye!
Take a look in youtube, there are some good tutorials there.https://github.com/openai/whisper Can someone make a tutorial video?
Whosper how do I srt the text file I translated?
Teşekkürler1 satırlık metin dosyası çıktısı yerine bir altyazı dosyası olan vtt dosyasını kullanırsınız, temelde bir srt ile aynıdır.
Bir srt'ye sahip olmakta ısrar ediyorsanız, vtt'yi hem vtt hem de srt'yi destekleyen bir altyazı oluşturma yazılımıyla açın ve onu bir srt olarak kaydedin, ancak bu gereksizdir, vtt daha iyi bir altyazı formatıdır.
It should still work with lower quality audio but the result will not be as good.Thanks
Does it work for videos with low sound quality?
OLD JAV VIDEO
I finally used it on Colab and it's amazing. I'm working through LULU-173 and BBAN-132 to post using it soon. Not a lot of cleaning up to do either! Weird thing is that I have one video, PED-020, and it stops reading at 26 minutes... which is exactly where pyTranscriber would stop.It should still work with lower quality audio but the result will not be as good.
soloporhoy666 is already lowering the quality to mp3s instead of using something like opus which would give better quality for the same size and is apparently still getting good enough results to get impressed.
Are you using the whisper+vad collab? because that version is the most stable.I finally used it on Colab and it's amazing. I'm working through LULU-173 and BBAN-132 to post using it soon. Not a lot of cleaning up to do either! Weird thing is that I have one video, PED-020, and it stops reading at 26 minutes... which is exactly where pyTranscriber would stop.
I did a test and got 1800 lines, it is possible that your source file is corrupted, to generate this test use the supjav version.I finally used it on Colab and it's amazing. I'm working through LULU-173 and BBAN-132 to post using it soon. Not a lot of cleaning up to do either! Weird thing is that I have one video, PED-020, and it stops reading at 26 minutes... which is exactly where pyTranscriber would stop.
Thanks for the tip! I used supjav, converted to a sound file via VLC, and ran it again. Made it to 1:07:00 this time - 1000 lines. Its strange because the converted file still has clear sound but Whisper/VAD process did nothing with it. I'll mess around with it again and let you know. I don't want to clutter up the jav subtitle posting area with my tech problems too much.I did a test and got 1800 lines, it is possible that your source file is corrupted, to generate this test use the supjav version.
View attachment 3126218