johnson2006christopher/adaptshot

Onboarding & discoverability: Windows execution-policy note, FAQ, a "your photographs and privacy" page, PyPI-safe README links, `keywords`/`Typing :: Typed`, social-preview meta, FUNDING.yml about 2 hours ago

documentation good first issue first-five-minutes

1

Few-shot image classification on an ordinary CPU, offline, with conformal prediction sets that say how much to trust each answer and a human-in-the-loop correction path. Torch-free ONNX inference, 120 MB, Python 3.10+.

Python

#computer-vision#conformal-prediction#cpu-inference#edge-ai#few-shot-learning#human-in-the-loop#offline-first#onnx#python#uncertainty-quantification

README, SECURITY.md, CONTRIBUTING, PR template and issue templates carry v0.2-era facts (Python 3.9+, "Current release v0.2.0", default backbone resnet18, 232 tests, 775 MB, four-stage gate, `ui/` in the tree) about 2 hours ago

documentation good first issue

johnson2006christopher/adaptshot

1

Few-shot image classification on an ordinary CPU, offline, with conformal prediction sets that say how much to trust each answer and a human-in-the-loop correction path. Torch-free ONNX inference, 120 MB, Python 3.10+.

Python

#computer-vision#conformal-prediction#cpu-inference#edge-ai#few-shot-learning#human-in-the-loop#offline-first#onnx#python#uncertainty-quantification

Errors: `BackboneError` tells users to install a non-existent `adaptshot[training]` extra; backbone problems are reported as `InvalidImageError`; config raises `ValueError` outside the `AdaptShotError` hierarchy about 2 hours ago

bug good first issue

johnson2006christopher/adaptshot

1

Few-shot image classification on an ordinary CPU, offline, with conformal prediction sets that say how much to trust each answer and a human-in-the-loop correction path. Torch-free ONNX inference, 120 MB, Python 3.10+.

Python

#computer-vision#conformal-prediction#cpu-inference#edge-ai#few-shot-learning#human-in-the-loop#offline-first#onnx#python#uncertainty-quantification

Open Issues Need Help