I am experiencing a terrible problems with installation of whisper-openai ai models. Basicly, all i wanted was initialisation of the model using the model = whisper.load_model("small") But any time the code starts processing, i get the same error: RuntimeError: Model has been downloaded but the SHA256 checksum does not not match. Please retry loading the model I have tried 99.99% of all of the solutions existing. Including cleaning the whisper cache, cleaning the pip cache, restarting the download many times, using 'git clone https://git.ffmpeg.org/ffmpeg.git ffmpeg' to install ffmpeg Then downloading the model manually and inserting it to the folder (i get the error like RuntimeError: some tensor operation failed because dim is missing ) Then i have tried installing model using huggingface + transformers, and i still get the same problem: whisper starts checking control sum error happens. What i have to do? I use MacBook based on M3 chip. I will appreciate any help. I have tried to install the whisper voice model, but got the SHA256 control sum error. Continue reading...
