# Probsparse Self-attention

**URL:** <https://discuss.huggingface.co/t/probsparse-self-attention/37806>\
**Category:** Beginners\
**Created:** [April 26, 2023, 2:29pm UTC](https://discuss.huggingface.co/t/probsparse-self-attention/37806 "2023-04-26T14:29:14Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Helectuss](https://avatars.discourse-cdn.com/v4/letter/h/87869e/32.png) [@Helectuss](https://discuss.huggingface.co/u/Helectuss)\
**Post date:** [April 26, 2023, 2:29pm UTC](https://discuss.huggingface.co/t/probsparse-self-attention/37806/1 "2023-04-26T14:29:14Z")

</div>

Hello,  
I was reading this blog post [Multivariate Probabilistic Time Series Forecasting with Informer](https://huggingface.co/blog/informer).  
At the end of probsparse attention it says “Please be aware that this is only a partial implementation of the `probsparse_attention`, and the full implementation can be found in 🤗 Transformers.”  
I cant seem to find the full implementation code in Transformers. Anyone can help me?

---

<div class="post-metadata">

**Author:** ![phimos](https://avatars.discourse-cdn.com/v4/letter/p/e68b1a/32.png) [@phimos](https://discuss.huggingface.co/u/phimos)\
**Post date:** [May 14, 2025, 2:05am UTC](https://discuss.huggingface.co/t/probsparse-self-attention/37806/2 "2025-05-14T02:05:25Z")

</div>

You can find it here:

> <https://github.com/huggingface/transformers/blob/b311a3f50697c9602cc5d13a5faf7f6059c392ca/src/transformers/models/informer/modeling_informer.py#L432>