# Useful compute\_metrics functions for perplexity

**URL:** <https://discuss.huggingface.co/t/useful-compute-metrics-functions-for-perplexity/23805>\
**Category:** 🤗Transformers\
**Created:** [September 29, 2022, 11:27pm UTC](https://discuss.huggingface.co/t/useful-compute-metrics-functions-for-perplexity/23805 "2022-09-29T23:27:05Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![jaedonvs](https://avatars.discourse-cdn.com/v4/letter/j/a9a28c/32.png) [@jaedonvs](https://discuss.huggingface.co/u/jaedonvs)\
**Post date:** [September 29, 2022, 11:27pm UTC](https://discuss.huggingface.co/t/useful-compute-metrics-functions-for-perplexity/23805/1 "2022-09-29T23:27:05Z")

</div>

I have trained a custom bert-base Transformer for MLM, and want to report the perplexity (opposed to eval\_loss) after each eval\_step. I have tried various different compute\_metrics functions to no avail. Any ideas?