# how to evaluate RAG-end2end

**URL:** <https://discuss.huggingface.co/t/how-to-evaluate-rag-end2end/7663>\
**Category:** Models\
**Created:** [June 29, 2021, 5:58pm UTC](https://discuss.huggingface.co/t/how-to-evaluate-rag-end2end/7663 "2021-06-29T17:58:35Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![giobin](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/giobin/32/28182_2.png) [@giobin](https://discuss.huggingface.co/u/giobin)\
**Post date:** [June 29, 2021, 5:58pm UTC](https://discuss.huggingface.co/t/how-to-evaluate-rag-end2end/7663/1 "2021-06-29T17:58:35Z")

</div>

Hi,

Is there a script like [eval\_rag.py](https://github.com/huggingface/transformers/blob/5257818e685c89fb473504afb5cbcd7b34301811/examples/research_projects/rag/eval_rag.py) but for RAG-end2end ? I see that the eval\_rag.py file in the RAG-end2end folder is the same as for RAG, but I am not sure on how to adapt it to load the finetuned dpr\_ctx\_encoder from a RAG-end2end checkpoint, which seems not that trivial.

Thank you