Feel free to use and enjoy
step1.py convert mp3 to srt step2.py convert srt to srt with 5s min_length and 15s max length to avoid too short/long sentences
Below is the version of software I have CUDA 12.6 use cmd and nvidia-smi to check your cuda version. refer to requirements.txt for other versions in case compatability.
python version 3.9.0 pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
Whisper is a general-purpose speech recognition model pip install -U openai-whisper ref: https://github.com/openai/whisper
pip install imageio pip install imageio[pyav] pip install imageio[ffmpeg]
There is a warning say 'weights_only=False' is not recommended. Below can fix. replace venvwhisper/Lib/site-packages/whisper/init.py with checkpoint = torch.load(fp, map_location=device, weights_only=True)