Lightnews — Scholar-powered news

Reproducing Uber's Alternating Direction Method of Multipliers (ADMM) based automated budget allocation system in JAX

gstechschulte.github.io/posts/2025-0...

Reproducing Uber's Marketplace Optimization

Uber allocates money across different regions and programs to incentivize riders and drivers to use Uber products. This incentive structure ultimately influences the market. This leads to the natural ...

gstechschulte.github.io

1 3

juanitorduz @juanitorduz.bsky.social · 14d

Amazing! Thank you !!!

2

juanitorduz @juanitorduz.bsky.social · 15d

Hehe classic.

juanitorduz @juanitorduz.bsky.social · 16d

💯

juanitorduz @juanitorduz.bsky.social · 20d

I went into the notebooks in detail and looks super interesting! I have never tried myself but now I’m very curious 🧐

1

juanitorduz @juanitorduz.bsky.social · 20d

Yay!

Reposted by juanitorduz

Jan Boelts @janboelts.bsky.social · 20d

Kudos to @sethaxen.com for implementing the Pyro wrapper that makes this possible (shipped in sbi v0.25)!

And thanks to @juanitorduz.bsky.social sharing the cookie factory example—it's a great accessible example for hierarchical inference.

Everything runs in Colab 📊

1 3

juanitorduz @juanitorduz.bsky.social · 20d

Cool! Any plans to add NumPyro support? 😄

1

Reposted by juanitorduz

Chris Fonnesbeck @fonnesbeck.bsky.social · 23d

A nice primer on normalizing flows by PyMC/PyTensor devs Ricardo and Jesse. pytensor.readthedocs.io/en/latest/ga...

Normalizing Flows in PyTensor — PyTensor dev documentation

pytensor.readthedocs.io

3 4

juanitorduz @juanitorduz.bsky.social · 25d

Here are the materials for the PyData Berlin 2025 talk on Stochastic Variational Inference with NumPyro:

- Slides: juanitorduz.github.io/html/intro_s...
- Notebook; juanitorduz.github.io/intro_svi/

Scaling Probabilistic Models with Variational Inference

juanitorduz.github.io

2

juanitorduz @juanitorduz.bsky.social · 25d

I think you can try with CustomDist. Here is an example discourse.pymc.io/t/creating-a...

Creating a LogStudentT distribution

Hi everyone, What is the suggested way to use a LogStudentT variable, a la LogNormal? Is there a more straightforward way than implementing a new Distribution? The use-case is the following. I’d lik...

discourse.pymc.io

1 1 4

juanitorduz @juanitorduz.bsky.social · Aug 31

If you need support from the (JAX) community I suggest having some “good first issues” on the repo :)

1 3

juanitorduz @juanitorduz.bsky.social · Aug 31

Great! There is some “related” work with JAX

- github.com/py-econometr...
- github.com/py-econometr...

2

juanitorduz @juanitorduz.bsky.social · Aug 23

Yes 100% (easy to iterate with LLMs)! I found many bugs in dat pre-processing steps.

juanitorduz @juanitorduz.bsky.social · Aug 20

Thanks for the resources! I had the same wish :)

juanitorduz @juanitorduz.bsky.social · Aug 20

Nice! I’ll join :)

1

juanitorduz @juanitorduz.bsky.social · Aug 15

Way (way) faster! Conda is, for all practical applications, infeasible. Try micromamba for example. Very similar api (you can keep the environment.yml) and faster :)

1 2

juanitorduz @juanitorduz.bsky.social · Aug 14

Or even micromamba mamba.readthedocs.io/en/latest/us...

Micromamba User Guide — documentation

mamba.readthedocs.io

1 2

juanitorduz @juanitorduz.bsky.social · Aug 5

I feel you, man, I really do.