Could you share your evaluation code for Retriever and Reranker?

[miaott1234]

Could you share your evaluation code for Retriever and Reranker?

[meghanadhpulivarthi]

I think the recall numbers reported of bge (recall@5: 0.5) are lower than what i got during my evaluation (recall@5: 0.65).

[mona-mahmoudi]

Could you share your evaluation code for Retriever and Reranker?

Hi @lesliepy99,

I am a Master’s student working on my thesis, and my research is closely related to your ICCAD 2024 paper:
"Customized Retrieval Augmented Generation and Benchmarking for EDA Tool Documentation QA".

I have explored your GitHub repository and really appreciate that the ORDQA benchmark and dataset are publicly available. I was wondering if it would be possible to share the full implementation code of your proposed RAG framework (including model training and inference scripts). Access to the complete code would greatly help me reproduce your results and build upon your approach for my thesis.

I assure you that the code will be used strictly for academic research purposes and will be properly cited.

Thank you very much for your time and consideration.