Skip to content

README.md

Latest commit

 

History

History
43 lines (29 loc) · 811 Bytes

README.md

File metadata and controls

43 lines (29 loc) · 811 Bytes

Model Auditor

A library for auditing ML models under distribution shifts.

Installation

Install directly from GitHub using pip:

pip install git+https://github.com/lukaskuhn/ModelAuditorCore.git

Or using uv:

uv pip install git+https://github.com/lukaskuhn/ModelAuditorCore.git

Usage

from model_auditor import ModelAuditor
from model_auditor.metrics import Accuracy, AUROC
from model_auditor.shifts import GaussianNoise, BrightnessShift

# Create auditor with your model
auditor = ModelAuditor(model)

# Add distribution shifts
auditor.add_shift(GaussianNoise(std=0.1))
auditor.add_shift(BrightnessShift(factor=0.2))

# Add metrics
auditor.add_metric(Accuracy())
auditor.add_metric(AUROC())

# Run audit
results = auditor.audit(dataset)

License

Apache-2.0