trinadhsriram02/ModelSentinel Security Scan

Scan PyTorch models for backdoors before deployment

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Stale
License
None
namedescriptionrequireddefault
model_pathPath to the model file to scanyes
risk_thresholdMaximum acceptable risk score (0-100)no40
num_classesNumber of output classes in the modelno10
fail_on_detectionFail the CI pipeline if backdoor is detectednotrue
namedescription
verdictScan verdict CLEAN SUSPICIOUS or BACKDOORED
risk_scoreRisk score 0-100
safe_to_deployWhether the model is safe to deploy