# Problem with KLqp inference for Poisson distribution

**URL:** <https://discourse.edwardlib.org/t/problem-with-klqp-inference-for-poisson-distribution/567>\
**Category:** General\
**Created:** [January 4, 2018, 6:56pm UTC](https://discourse.edwardlib.org/t/problem-with-klqp-inference-for-poisson-distribution/567 "2018-01-04T18:56:05Z")\
**Posts on this page:** 1\
**Showing post:** 2

<div class="post-metadata">

**Author:** ![pedrofale](https://yyz1.discourse-cdn.com/flex035/user_avatar/discourse.edwardlib.org/pedrofale/32/219_2.png) [@pedrofale](https://discourse.edwardlib.org/u/pedrofale)\
**Post date:** [February 5, 2018, 11:47am UTC](https://discourse.edwardlib.org/t/problem-with-klqp-inference-for-poisson-distribution/567/2 "2018-02-05T11:47:42Z")

</div>

Hi! I am having the exact same problem. Have you solved it? @dustin has said a lot of times in this forum that `ed.KLqp` doesn’t work well with Gamma and Dirichlet variational approximations (eg: [Error in inference.run() for a mixture model](https://discourse.edwardlib.org/t/error-in-inference-run-for-a-mixture-model/175/8?u=pedrofale)), but I never found an explanation for this.

---

_[View the full topic](https://discourse.edwardlib.org/t/problem-with-klqp-inference-for-poisson-distribution/567)._