# Stochastic (L-)BFGS

**URL:** <https://discourse.julialang.org/t/stochastic-l-bfgs/93138>\
**Category:** Optimization (Mathematical)\
**Tags:** question\
**Created:** [January 18, 2023, 9:13am UTC](https://discourse.julialang.org/t/stochastic-l-bfgs/93138 "2023-01-18T09:13:27Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![jbrea](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jbrea/32/3879_2.png) [@jbrea](https://discourse.julialang.org/u/jbrea)\
**Post date:** [January 18, 2023, 9:13am UTC](https://discourse.julialang.org/t/stochastic-l-bfgs/93138/1 "2023-01-18T09:13:27Z")

</div>

Anyone knows of a julia implementation of a stochastic (L-)BFGS variant, like e.g. [described here](http://proceedings.mlr.press/v51/moritz16.html) or [described here](https://epubs.siam.org/doi/abs/10.1137/140954362) or [implemented in pytorch here](https://github.com/nlesc-dirac/pytorch)?

---

<div class="post-metadata">

**Author:** ![baggepinnen](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/baggepinnen/32/693_2.png) [@baggepinnen](https://discourse.julialang.org/u/baggepinnen)\
**Post date:** [January 18, 2023, 4:33pm UTC](https://discourse.julialang.org/t/stochastic-l-bfgs/93138/2 "2023-01-18T16:33:26Z")

</div>

[GitHub - baggepinnen/FluxOptTools.jl: Use Optim to train Flux models and visualize loss landscapes](https://github.com/baggepinnen/FluxOptTools.jl) has an implementation, but I haven’t tested it much.