summarity

9 ms·

121.
by summarity 2y ago
I was working with the "ego." programs in particular: https://www.ib-sachsen-anhalt.de/gruender/gruenden-in-sachse... This might be different from state to state. There are also EU grants you can apply for, w
122.
by summarity 2y ago
Here's a model that exists in Germany, which I like: You can present a business plan to the state's investment bank and apply for several financial aides, including: * 1.5 years of universal basic income for you plus up to 2 other
123.
Eagle-3 Speculative Decoding for LLM Inference (5.6x speedup)(github.com)
2 points by summarity 2y ago|0 comments
124.
by summarity 2y ago
One of my teams at GitHub develops Copilot Autofix, which suggests fixes based on CodeQL alerts (another of my teams’ projects). Based on data of actual devs interacting with alerts and fixes, we see an average 3x speed up in time to fix ov
125.
by summarity 2y ago
You can chain multiple Mac Studios using exo for inference, you'd "only" need two of these. There's a bottleneck in the RMA speed over TB5, but this may not matter as much for a MoE model.
126.
The Picat Programming Language(picat-lang.org)
2 points by summarity 2y ago|0 comments
127.
by summarity 2y ago
https://litesync.io/en/ is a mature way of distributing SQLite with arbitrary topologies and no modifications to client code. Supports multi primary write and DDL replication. It’s a paid solution, but that follows the
128.
by summarity 2y ago
Plus one for Kusto. I’ve never had this much success teaching people data analysis. It’s good for simple obvious and powerful composed queries. Though I work at GitHub so still too close to the mothership for an unbiased opinion. There’s so
129.
by summarity 2y ago
Oh very neat. I had worked with Fennel inside touchOSC to create custom control surfaces and used it extensively for chord identification logic. The code is private, but here's an excerpt (e.g. norm-chord tries to identify chords based
130.
by summarity 2y ago
Strongly opinionated languages beget strong opinions on the same, both positive and negative.
131.
PackCC – Compact PEG Parser Generator for C(github.com)
1 points by summarity 2y ago|0 comments
132.
by summarity 2y ago
Very suspiciously similar comments here
133.
by summarity 2y ago
Also CLIPS has an entire book on creating an interactive fiction game (Adventures in Rule based Programming)
134.
by summarity 2y ago
As someone whose day job involves a lot of graph analysis and logic programming[0], I'm always excited to see new applied research in this area. More energy is needed here. Logic systems will be a key part of solving problems of hybrid
135.
by summarity 2y ago
> creation of self in the age of social media That's a large part of the field of Cyberpsychology, and of course there's quite a body of research already.
136.
Gdelt – A Global Database of Society(gdeltproject.org)
4 points by summarity 2y ago|0 comments
137.
by summarity 2y ago
The template doesn't match Unsloth's recommendation: https://news.ycombinator.com/item?id=42662106
138.
by summarity 2y ago
Does it include the unsloth fixes yet?
139.
by summarity 2y ago
Can't mention VB6 without one of the biggest archives of source-available software for it: https://github.com/Planet-Source-Code On the topic of "wait what that's VB6", here's an entire 3D modeling
140.
Microsoft releases TRELLIS dataset: 500k 3D assets for model generation training(github.com)
2 points by summarity 2y ago|0 comments
141.
by summarity 2y ago
Pretty amazing. Scale consistency is an issue looking at the scene examples, but for one-off or background assets, pretty neat!
142.
XBOW AI autonomously discovers path traversal in EDDI (CVE-2024-53844)(xbow.com)
2 points by summarity 2y ago|0 comments
143.
by summarity 2y ago
> Just tokenizing based on a few common characteristics (brace pairing, quotes, indentation, newlines, etc.) should let you trim a ton without knowing anything about the language, I imagine? Yep, here's an explanation from a related
144.
by summarity 2y ago
That is what they (explicitly) said in the email: users using next to no bandwidth were offsetting the costs for the heaviest users. That was no longer sustainable.
145.
by summarity 2y ago
Yes, that was it! I used the same hack in 2012. Fun story: as a prize for winning the science fair with my robot arm, I got to go to Munich to visit ASDEX Upgrade (fusion experiment). I was staying there for a week and just spent each day w
146.
by summarity 2y ago
I also recommend this video as a resource on low cost EEG from someone who tried living with it: https://youtu.be/c-LyXBCNsrA?si=7xoM1OGTNRbMsO_P (skim through) That goes over several nicer multichannel options. The NeuroSk
147.
by summarity 2y ago
I was in high school and this was for a science fair. It kind of worked, you could more or less reliably move one axis of the arm at a time. I directly used the relax/focus metrics and relied on a training phase where the participant w
148.
by summarity 2y ago
O hey, I have also done this before around 10 years ago. Initially my project was also based on a hacked Mindflex (it was a robot arm). I then reached out to NeuroSky and they just sent me a bunch of their higher tier dev kits for free. No
149.
by summarity 2y ago
It’s true. DeepL is based in part on the linguee data, which has long been the highest quality real world language use database. It excels at not translating literally, but actual meaning. While not being impaired by also having to be a gen
150.
by summarity 2y ago
I can't get past the sign-up form: "503 Service Temporarily Unavailable"

More ›