The README notes regulation evaluation as ``` cd regulation # Attack evaluation on regulatory compliance python evaluate_attack.py --model_id OpenSafetyLab/MD-Judge-v0_2-internlm2_7b # Conversation evaluation python evaluate_conversation.py --model_id meta-llama/Llama-Guard-3-8B # Query evaluation python evaluate_query.py --model_id allenai/wildguard ``` The `evaluate_attack.py` file in turn tries to access `dataset_path = f"{base_dir}/dataset/regulation/gdpr/query/unsafe.jsonl"` However, there is no such file.
Activity