I started to make a pretty big refactoring of the code base, if anyone is interested:
https://github.com/srad/QualityScaler/tree/resolution-fit
Changes I made:
- refactored the
QualityScaler.py file into individual modules and classes
- No global variables (only file with default values)
- and already fixed minor issue and added new features, like
- showing only the available GPU by name in the GPU dropdown box
- automatically setting the maximal available GPU memory when the GPU is selected (and on launch)
- logging instead of printf
- Using correct python and patterns naming convention.
- Adding more type hints
- The entire codebase is already 500 lines less code than the original
QualityScaler.py file (cloc . --exclude-dir=.venv,venv,AI-onnx,Assets --exclude-ext=md,txt,bat,xml) and it will shrink more
It not entire fixed yet, since I'm still working on it. But the application can be launched and everything looks right.
I started to make a pretty big refactoring of the code base, if anyone is interested:
https://github.com/srad/QualityScaler/tree/resolution-fit
Changes I made:
QualityScaler.pyfile into individual modules and classesQualityScaler.pyfile (cloc . --exclude-dir=.venv,venv,AI-onnx,Assets --exclude-ext=md,txt,bat,xml) and it will shrink moreIt not entire fixed yet, since I'm still working on it. But the application can be launched and everything looks right.