# Why not always use Adam optimizer

**URL:** <https://community.deeplearning.ai/t/why-not-always-use-adam-optimizer/245823>\
**Category:** Structuring Machine Learning Projects\
**Tags:** coursera-platform\
**Created:** [December 22, 2022, 1:37pm UTC](https://community.deeplearning.ai/t/why-not-always-use-adam-optimizer/245823 "2022-12-22T13:37:11Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![Christian\_Simonis](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/christian_simonis/32/142583_2.png) [@Christian\_Simonis](https://community.deeplearning.ai/u/Christian_Simonis)\
**Post date:** [December 23, 2022, 10:04pm UTC](https://community.deeplearning.ai/t/why-not-always-use-adam-optimizer/245823/4 "2022-12-23T22:04:08Z")

</div>

In addition to 2)  
If Adam does not converge well, AMSGrad might be worth a look, see also:  
[https://johnchenresearch.github.io/demon/](https://johnchenresearch.github.io/demon/)

Here also some other algorithms are explained like QHM [(Quasi-Hyperbolic Momentum)](https://johnchenresearch.github.io/demon/#fn:6) which decouples the momentum term from the current gradient when updating the weights which can also be beneficial!

Best regards  
Christian

---

_[View the full topic](https://community.deeplearning.ai/t/why-not-always-use-adam-optimizer/245823)._