I try to use whisper at this link: https://colab.research.google.com/g...in/WhisperWithVAD.ipynb#scrollTo=teF-Ut8Z7Gjp
My settings is No translation, vad 0.3 and chunk 1-1.1. With this chunk value, the output contains many short sentences rather than few long sentences. I think in JAV, they don't talk long, so we should go with many short sentences.
After getting SRT file, I will translate it with Google Chrome. I believe Google translate better than DeepL or other services.
I also have write a small program to clean up SRT file after translation. It fixes some minor issues. Sometimes the ID of subtitle is translated to "five", "six" instead of remaining as number, the empty seperator line is missing (cause 2 lines to be displayed as 1), remove useless lines such as "ah", "oh", etc.
Here's some subtitles I made using above method. I don't bother editting them because they're good enough for me.
XRW-121
ATID-263
CMN-142
CMV-055
DDK-208
DNIG-002
HAVD-064
SDMS-474
SSPD-030
My settings is No translation, vad 0.3 and chunk 1-1.1. With this chunk value, the output contains many short sentences rather than few long sentences. I think in JAV, they don't talk long, so we should go with many short sentences.
After getting SRT file, I will translate it with Google Chrome. I believe Google translate better than DeepL or other services.
I also have write a small program to clean up SRT file after translation. It fixes some minor issues. Sometimes the ID of subtitle is translated to "five", "six" instead of remaining as number, the empty seperator line is missing (cause 2 lines to be displayed as 1), remove useless lines such as "ah", "oh", etc.
Here's some subtitles I made using above method. I don't bother editting them because they're good enough for me.
XRW-121
ATID-263
CMN-142
CMV-055
DDK-208
DNIG-002
HAVD-064
SDMS-474
SSPD-030