GuardBench is a Python library for evaluating guardrail models that detect unsafe content in...

Tokens:19,847
Snippets:216
Trust Score:7.7
Update:2 months ago
Tokens:
Raw