We introduce SFR-RAG, a 9B LLM trained with an emphasis in contextual comprehension and retrieval augmented generation (RAG) use case.
We also introduce ContextualBench - a compilation of 7 popular contextual question answering benchmarks to evaluate LLMs in RAG application.
This code is licensed under CC-BY-NC 4.0. This is a research only project.
Gemma is provided under and subject to the Gemma Terms of Use found at ai.google.dev/gemma/terms.
Users need to make their own assessment regarding any obligations or responsibilities under the corresponding licenses or terms and conditions pertaining to the original datasets and data.
We want to acknowledge the work which have made contributions to our paper and the agent research community! If you find our work useful, please consider to cite
@article{nguyen2024sfrrag,
title={SFR-RAG: Towards Contextually Faithful LLMs},
author={Nguyen, Xuan-Phi and Pandit, Shrey and Purushwalkam, Senthil and Xu, Austin and Chen, Hailin and Ming, Yifei and Ke, Zixuan and Savarese, Silvio and Xong, Caiming and Joty, Shafiq},
year={2024}
}