# Hamiltonian Monte Carlo gets stuck at tiny local maxima

**URL:** <https://discourse.julialang.org/t/hamiltonian-monte-carlo-gets-stuck-at-tiny-local-maxima/108389>\
**Category:** Probabilistic Programming\
**Tags:** question\
**Created:** [January 5, 2024, 2:18pm UTC](https://discourse.julialang.org/t/hamiltonian-monte-carlo-gets-stuck-at-tiny-local-maxima/108389 "2024-01-05T14:18:34Z")\
**Posts on this page:** 1\
**Showing post:** 14

<div class="post-metadata">

**Author:** ![torfjelde](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/torfjelde/32/206542_2.png) [@torfjelde](https://discourse.julialang.org/u/torfjelde)\
**Post date:** [January 12, 2024, 3:01pm UTC](https://discourse.julialang.org/t/hamiltonian-monte-carlo-gets-stuck-at-tiny-local-maxima/108389/14 "2024-01-12T15:01:33Z")

</div>

> [@marcsgil](#):
>
> Is that tempering thing that is mentioned in your link already implemented in Julia?

Just a heads up, this is in the works at [MCMCTempering.jl](https://github.com/TuringLang/MCMCTempering.jl). I had to put it on hold for a bit, but am about to return to it. Unfortunately it’s not _quite_ ready for “public consumption” just yet, but it will be in not too long:)

And I’d expect tempering to be quite useful in this scenario!

There is also [Pigeons.jl](https://github.com/Julia-Tempering/Pigeons.jl) which is _slightly_ more restrictive (though has fantastic support for distributed workloads), that should also work on Turing.jl: [Turing.jl model · Pigeons.jl](http://pigeons.run/dev/input-turing/). Maybe give this a try?🙂

---

_[View the full topic](https://discourse.julialang.org/t/hamiltonian-monte-carlo-gets-stuck-at-tiny-local-maxima/108389)._