thenipper

44 ms·

91.
by thenipper 4y ago
There is a ton of interesting work being done with solvers adjacent to machine learning. I just finished a job where I was working with an operations research team solving inventory placement and transportation
92.
by thenipper 4y ago
This reminds me of the old shadow run rpg books. They had in character comments throughout the books. There were whole subplots and narrative arcs just in the comments. It was great.
93.
by thenipper 4y ago
Pyinfra supports home brew: https://docs.pyinfra.com/en/latest/operations/brew.html I’ve thought about using it to set up my machines/machines at work but haven’t gone down that road yet.
94.
by thenipper 4y ago
Most legacy ML models I’ve had to move to production aka not on a laptop. Oh you meant in a good way…
95.
by thenipper 4y ago
Yeah Cambridge brewing company has made a nice knotweed beer
96.
by thenipper 4y ago
Ha. I have the opposite reaction.
97.
by thenipper 4y ago
As the other comment said you can do gates. You can also write lua mods. I’ve done that for some stats/dashboarding.
98.
by thenipper 4y ago
Check out plumbum: https://plumbum.readthedocs.io/en/latest/ It supports mac, Linux and windows.
99.
by thenipper 4y ago
I love python. I’m pretty good at it or so I like to think. But yet circular imports trip me up constantly.
100.
by thenipper 4y ago
It adds some overhead but half the time I’m lazy and just use Ray: https://www.ray.io/ to handle my concurrency for me. That being I agree with you some better abstractions would be nice.
101.
by thenipper 4y ago
This is pretty cool! Up until about a month ago I had been doing engineering for OR models for a few different companies. It’s neat to see something in this space compared to “regular” ML model serving
102.
by thenipper 4y ago
I use an M1 MacBook pro for work. I’ve had no problems in general. There are few wrinkles with python development. The main thing is about various data specific libraries getting ported over. Most of the problems are solved with a quick goo
103.
by thenipper 4y ago
To be fair, that’s a kind of amazing screw up. Like… do they even have gherkins at McDonalds?
104.
by thenipper 4y ago
I just came back from a family vacation. I looked at getting an Uber or Lyft from airport home. They were all like a twenty minutes wait and 50 bucks. We took a cab that was already there and it came in at about 40 bucks.
105.
by thenipper 4y ago
Also there is optimization/operations research if you want something a bit different. Less buzzwordy too.
106.
by thenipper 4y ago
I’ve got that problem too with my username minus the “the” at gmail. It’s so annoying.
107.
by thenipper 4y ago
This! I went from being a hobbyist dev -> leading a team at a large tech company. The biggest learning I had to do was around writing code with others. It’s one thing to be good at writing code/making a project. It’s another to writ
108.
by thenipper 5y ago
Nope. It was a crazy stored procedure. When all you know is SQL… you’ll solve some crazy problems using sql. Oof. That tbh sounds worse
109.
by thenipper 5y ago
I once refactored a 15k plus line sql file. Fun times
110.
by thenipper 5y ago
I wonder if this is also to make it harder to recruit folks who work at Meta. If everyone’s title is the same external recruiters don’t know who to target by level.
111.
by thenipper 5y ago
I work with operations research teams in a blended model of engineering being embedded with the OR Scientists. I really prefer it. Code can get to prod a lot quicker and we don’t have the “throw it over the fence to engineering” issues that
112.
by thenipper 5y ago
Is the second one wayfair? It echoes what all my friends there are saying.
113.
by thenipper 5y ago
This comment is dead on. I’m SWE who builds data/optimization systems and someone with this skill set would be a great hire. Even if you wanted to get totally away from data most any modern system is going to need to interface with dat
114.
by thenipper 5y ago
Almost. In my experience those periods of employment are just blank.
115.
by thenipper 5y ago
We've been thinking about trying this out at work for some of our data pipelines/simplified models. The speed/ergonomics look great.
116.
by thenipper 5y ago
I’ll totally second this. It takes away a lot of the tedium of multiprocessing programming. I implemented it at my last job and it was a game changer. Simplified a lot of gnarly cluster management and let me focus on my code rather then fig
117.
by thenipper 5y ago
I’ve got two step kids who play roblox/Minecraft obsessively. The network effect in playing what their friends play is super strong. That combined with as busy parent there is a reluctance to switch to anything else. They don’t want to
118.
by thenipper 5y ago
Related I learned to just get oral rehydration salts when I worked at an NGO. Good for hydration on a hot day. But also an amazing hangover cure.
119.
by thenipper 6y ago
You can use Prophet in R for sure
120.
Fastcore: A library that extends Python with new features(fastpages.fast.ai)
187 points by thenipper 6y ago|91 comments

More ›