Philipp Schiele

PhD student working on optimization

[Philipp Schiele @phschiele1]

Amazing work being done at @marimo_io!

[marimo @marimo_io]

We're excited to announce that @Github has selected marimo to participate in its accelerator for open source AI projects! 🎉

We're reinventing the Python notebook for modern AI/ML teams, and we've got lots to share in the coming weeks. Stay tuned!

https://t.co/o9RXBR9sGE

[marimo @marimo_io]

We're excited to announce that @Github has selected marimo to participate in its accelerator for open source AI projects! 🎉

We're reinventing the Python notebook for modern AI/ML teams, and we've got lots to share in the coming weeks. Stay tuned!

https://t.co/o9RXBR9sGE

[marimo @marimo_io]

If you agree — pip install marimo && marimo tutorial intro.

[myles @themylesfiles]

localhost:2718 for @marimo_io was not random as well

[Philipp Schiele @phschiele1]

Much of our software relies on optimized libraries implementing BLAS & LAPACK. Riley Murray's work on randomized numerical algorithms is pushing these boundaries further. Check out the video for a deep dive.

https://t.co/ZoqBMto6FS

[CVXPY @cvxpy_team]

🌟 CVXPY just surpassed 5,000 stars on GitHub! 🌟

Thanks to all the contributors and users who made this possible!

🔗 CVXPY on GitHub
https://t.co/wIaKZat8lu

📢 Join our Discord community to stay connected
https://t.co/cjK0nyzDEL

[Fabian Schaipp @FSchaipp]

What are nice applications for constrained #optimization problems, where objective and/or constraints are *already trained* neural networks (more generally, mappings that only allow for eval and jacobians)?

[arXiv math.OC Optimization and Control @mathOCb]

Vinit Ranjan, Bartolomeo Stellato: Verification of First-Order Methods for Parametric Quadratic Optimization https://t.co/oWmele4tWt https://t.co/HwpPEZl99d

[Philipp Schiele @phschiele1]

The cherry on top: you can give it a try right in your browser without the need for a local install, powered by WebAssembly:
https://t.co/RXwhSnAWOU 8/8

[Philipp Schiele @phschiele1]

📢 PSA: If you are working with Python notebooks, do yourself a favor and check out marimo. 🧵1/8

[Philipp Schiele @phschiele1]

Check out their GitHub here: https://t.co/w7sEb3MGPY ⭐
#python #notebooks #datascience 7/8

[Bartolomeo Stellato @b_stellato]

📢📢📢@Princeton Workshop on Optimization, Learning, and Control #OLC24, June 27-28, 2024 🌐https://t.co/wMNCw8zdnS

Poster 📈 submission deadline extended to March 10!!!🗓️

#optimization #learning #ml #control #orms

@NSF @Princeton ORFE @EPrinceton @gurobi @PrincetonSML @IBM

[Akshay Agrawal @akshaykagrawal]

Stephen Boyd, on the practical value of a training in math, with lessons for building software.

"In my opinion, the real value of math in applied settings, ... is it gives you clarity of thought.

Here we are in Silicon Valley ... I could get one team, and say, 'go build this thing.' With a lot of high-level tools now, you can bang something together that looks like it's working ...

But there's another style that's just beautiful. People who think about it very deeply, often trained in math or something abstract, they don't do anything for a day or two. [Then] they look up and go, 'The 87 things we've been asked to implement are actually all instances of only 3 different things. So we're going to implement [those 3 things] correctly.'

When people just hack something together to knock off 87 requirements, it's going to be horrible code, badly structured, ... it won't be correct ... you definitely won't be able to extend it.

Whereas when you see people who think abstractly, and took the time to work out the correct abstractions, you get beautiful lean code that has very high probability of being correct. As a purely practical matter, the cost of ownership of that project is a lot less because it will work, and it's extensible and maintainable. It's like Linux, for example. That's a good model for this."

[Akshay Agrawal @akshaykagrawal]

Kinda awesome ...! Helps put in perspective just how much we've shipped these past few months.

Also — throwback to late nights in college spent cramming an 8.5x11 with as many formulas as possible ...

[marimo @marimo_io]

We published our first blog post! In it, we explain why we started marimo, and the problems it solves:

https://t.co/YoAdhObBpf