MM-GRADE: A Multi-Modal EDA Tool Documentation QA Framework Leveraging Retrieval Augmented Generation
- becnhmark (The ORD-MMBnech fiels)
- q_images (The screenshot images for the evaluation benchmark)
- openroad_document_chunts.json (The Openroad documentation chunks)
- QA.jsonl (The 120 query-document-answer triplets in ORD-MMBench)
- evaluation
- ORD-MMBench-Scoring-criteria.xlsx (The scoring criteria for each query in ORD-MMBench)
- answers/ (The generated answers of all the evaluated RAG flows on ORD-MMBench)
- scores/ (The LLM-scores of all the evaluated RAG flows on ORD-MMBench)
- training_dataset (The training dataset for the multi-modal retriever model and the generator model)