Datenstrom

10 ms·

91.
by Datenstrom 6y ago
I had a similar idea when I was working with them but I never got around to trying it. I was primarily using them to compose other behaviors I had developed. It is not my specialization so I'm not up on the latest research around them
92.
by Datenstrom 6y ago
I do not believe that is true. Pure FSMs are not Turing complete[1] while behavior trees can be made trivially so. BTs are not isomorphic to FSMs. I am not aware of any system that can be modeled in an FSM and not in a behavior tree. [1]:
93.
by Datenstrom 6y ago
Found this to be a great introduction to an alternative way to model behaviors than a FSM. I used behavior trees for robotics and AI applications and found them generally simpler and that they produced much more modular and reusable code.
94.
Behavior Trees in Robotics and AI(arxiv.org)
36 points by Datenstrom 6y ago|8 comments
95.
by Datenstrom 6y ago
When I worked on robotics I often found myself reaching for behavior trees for anything complex over a FSM. See section 2 of "Behavior Trees in Robotics and AI"[1] for an example of how much simpler they are. I think they are popu
96.
by Datenstrom 6y ago
Unfortunately someone is going to wrap up super-resolution for critical tasks and sell it likely causing many people harm or at least inconvenience. I have already tried to talk some companies out of using it for police/surveillance ty
97.
by Datenstrom 6y ago
Krydus | Backend, Frontend, ML & ML Ops Engineers | Full-Time | Remote (USA) At Krydus we are developing tools to accelerate and automate the boring parts of a ML Engineers job, going from raw unstructured data to a well balanced datase
98.
by Datenstrom 6y ago
Krydus | Backend, Frontend, ML & ML Ops Engineers | Full-Time and Part-Time | Remote | Must be a US Citizen At Krydus we are developing tools to accelerate and automate the boring parts of a ML Engineers job, going from raw unstructured
99.
by Datenstrom 6y ago
Glad to see Podman here, worked with it at a past job where services were deployed on RedHat. Converting all existing deployment scripts to podman was as easy as do a recursive search and replace `s/docker/podman/` and there
100.
by Datenstrom 6y ago
What does this offer that cortex[1] does not? With cortex supporting scale to zero it costs me less than $1/mo to idle and autoscales to a cap only when I need it even using spot instances when available if it is configured. Disclaimer
101.
by Datenstrom 6y ago
> Not to mention that you need an extremely loose definition of "copy". Makes me wonder, with a nearly perfect simulation for a reinforcement learning agent and a human brain or equivalent architecture, would you be able to mak
102.
by Datenstrom 6y ago
My vim configuration, environment, and screen splitting using tmux giving me CLI tools in other windows together gives me every single feature you listed and more. My shell and environment being Turing complete not counting vimscript I imag
103.
by Datenstrom 6y ago
Krydus | Backend, Embedded, Ops and ML Engineers | Full-Time and Part-Time | Remote | US Preferred We are a fast growing artificial intelligence startup building a computer vision product that quickly provides data insights and cleaning giv
104.
by Datenstrom 6y ago
> manufacturers don't care about Linux support This is why I have been using System76 personally for the last 5 years, since I first found out about them when given one by an employer. The first linux box I have had that didn't
105.
by Datenstrom 6y ago
DoH will just be the start of another arms race. If you would like to control DoH also user LinuxBender responded to a question I asked about DoH blocklists with these repos in another post. I haven't got around to trying it out yet th
106.
OpenWISP: A Hackable Network Management System(openwisp.org)
11 points by Datenstrom 6y ago|1 comments
107.
by Datenstrom 6y ago
Is there a comprehensive list to block somewhere? I would like to do the same.
108.
by Datenstrom 6y ago
Yes, had this problem when I was setting up my pihole and took me a bit to figure out why it wasn't working with firefox. Imagine the sysadmin nightmare if you had to configure 100 different applications on every one of your systems to
109.
by Datenstrom 6y ago
Otherwise I would definitely not say they tried to micromanage everything. Maybe they thought they had a good understanding of that particular problem I am not sure, but the solution had already been pitched to the customer before I transfe
110.
by Datenstrom 6y ago
> When we notified management they replied, “We’re in a build the plane while flying situation, how can we get this out now without doing a big rewrite?” Even worse I tried to catch this before it happened at one company and was stopped.
111.
AMD ROCm Platform(rocmdocs.amd.com)
3 points by Datenstrom 6y ago|0 comments
112.
by Datenstrom 6y ago
With the new generation of cards from Nvidia being significantly worse for GPGPU scientific compute I hope AMD moves into that market and is truly an Nvidia killer. The new Nvidia cards not only consume far more power and space making 4-GPU
113.
by Datenstrom 6y ago
Yes it is possible to make those changes and even if it wasn't it does not detract from the impressive and useful innovations Julia has made. That is why I usually refer to them as cosmetic but that is possibly not the correct word eit
114.
by Datenstrom 6y ago
There is an immense value in most of the ecosystem converging on Python over the last decade and especially with leaving MATLAB behind in the dust. I don't think decade long transitions from old tech to new tech is a bad time scale or
115.
by Datenstrom 6y ago
Yes, I have done non-trivial implementations of a number of SoTA models in Julia. The framework I've used is Flux[1] which I love for it's simplicity, it is very much like the DarkNet[2] framework in that regard which is refreshin
116.
by Datenstrom 6y ago
True, I always took math notes in LaTeX which I am fluent in and still it is much slower. Makes you think about how you can distill an idea down into key concepts which leads to active learning IMO. I still keep all my math notes in LaTeX a
117.
by Datenstrom 6y ago
Krydus | Various Engineers and Scientists | Full-Time | REMOTE & ONSITE (depending on position) | DC & Seattle We are a bootstrapped employee owned company. Our primary product is focused on applying active learning to provide data
118.
by Datenstrom 6y ago
These are pretty good results, but not SOTA currently EfficientNet-B7 outperforms this on ImageNet-1000. ImageNet in particular is around 84% top-1 and 97% top-5. Claiming the last few percentages can be much harder than the first so it wou
119.
End-to-End Encryption and Internet Privacy Solutions(cyph.com)
4 points by Datenstrom 6y ago|0 comments
120.
Project Ultramarine: The Sun Powered Sensor Box(ultramarine.cloud)
2 points by Datenstrom 6y ago|1 comments

More ›