# Evaluating RAG only with open-source

**URL:** <https://discuss.huggingface.co/t/evaluating-rag-only-with-open-source/77057>\
**Category:** Intermediate\
**Created:** [March 12, 2024, 10:02am UTC](https://discuss.huggingface.co/t/evaluating-rag-only-with-open-source/77057 "2024-03-12T10:02:22Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![dkoterwa](https://avatars.discourse-cdn.com/v4/letter/d/e9c0ed/32.png) [@dkoterwa](https://discuss.huggingface.co/u/dkoterwa)\
**Post date:** [March 12, 2024, 10:02am UTC](https://discuss.huggingface.co/t/evaluating-rag-only-with-open-source/77057/1 "2024-03-12T10:02:22Z")

</div>

Hi,  
are there any RAG evaluation tools based solely on open-source models (for example using only HuggingFace models)? I would like to load models for generating responses/judging responses from HuggingFace and compute metrics based on the scores of the judge.

---

<div class="post-metadata">

**Author:** ![elfailali](https://avatars.discourse-cdn.com/v4/letter/e/c67d28/32.png) [@elfailali](https://discuss.huggingface.co/u/elfailali)\
**Post date:** [May 24, 2024, 8:53pm UTC](https://discuss.huggingface.co/t/evaluating-rag-only-with-open-source/77057/2 "2024-05-24T20:53:30Z")

</div>

Try Ragas framework