Hi — we're integrating GRANNY into a fruit-QC platform (via pip install granny) and ran into a docs/release mismatch: the readthedocs documentation describes the adjustable parameters (--confidence, --starch_threshold, etc.) from v1.1.0, but PyPI's newest version is still 1.0a1 (July 2024), where those parameters don't exist yet — e.g. Segmentation._segmentInstances calls predict() without conf, so YOLO's 0.25 default can't be changed.
Since v1.1.0 is already tagged and released here on GitHub, would you be able to publish it to PyPI? Happy to provide any testing feedback in return — we're actively running the segmentation + starch pipeline in production against orchard tray photos.
Hi — we're integrating GRANNY into a fruit-QC platform (via pip install granny) and ran into a docs/release mismatch: the readthedocs documentation describes the adjustable parameters (--confidence, --starch_threshold, etc.) from v1.1.0, but PyPI's newest version is still 1.0a1 (July 2024), where those parameters don't exist yet — e.g. Segmentation._segmentInstances calls predict() without conf, so YOLO's 0.25 default can't be changed.
Since v1.1.0 is already tagged and released here on GitHub, would you be able to publish it to PyPI? Happy to provide any testing feedback in return — we're actively running the segmentation + starch pipeline in production against orchard tray photos.