thuttinger
Joined 4 years ago
Joined 4 years ago
It's what I want to try next. Written in go, it looks promising.
Small improvement: the node script could check if the model files are already present at the download location and not download them again in this case.
On Mac OS, run for example:
ncdu -x --exclude /Volumes --exclude /System/Volumes /
This scans the root filesystem but excludes the Volume mounts specifically (the -x option to limit the scan to the current filesystem doesn't work properly on mac os).
Navigate through the folders, sorted by size, from there. Press d to delete a folder.If you want to give it a shot, you can find the python script in this repo: https://github.com/tobiashuttinger/openai-whisper-realtime
A bit more context on how it works: The systems default audio input is captured with python, split into small chunks and is then fed to OpenAI's original transcription function. It tries (currently rather poorly) to detect word breaks and doesn't split the audio buffer in those cases. With how the model is designed, it doesn't make the most sense to do this, but i found it would be worth trying. It works acceptably well.
- Read an .icc file from disk
- parsed the file and extracted the VCGT (video card gamma table)
- wrote the VCGT to the video card for a specified display via amdgpu driver APIs
The only thing I had to fix was the ICC parsing, where it would parse header strings in the wrong byte-order (they are big-endian).