-
I was able to package the onnx models in Electron so that when installed (at least via AppImage), the models actually load on the target machine and transcriptions work. I had a lot of doubts about whether this would actually work.
A log of progress on a podcast assistant app I'm making.