efskap

10 ms·

121.
by efskap 1y ago
> 14.03.02 - Bolts; Fasteners, bolts; Fasteners, nails; Fasteners, screws; Nails (hardware); Nuts (hardware); Rivets; Screws; Tacks, thumb; Thumbtacks > 26.01.03 - Circles, incomplete (more than semi-circles); Incomplete circles (more
122.
by efskap 1y ago
I'm very interested as an e-ink enjoyer, as the most capable tablets run Android, mainly the Boox ones. Sure there's the PineNote that runs ordinary Debian but I hear it's not quite there yet usability-wise, and it doesn&#x2
123.
by efskap 1y ago
>which is no RAM management at all I'm sure the Android one is much more aggressive, but Linux's OOM killer isn't too different is it?
124.
by efskap 1y ago
Very cool, like <input type="password"> except with whitespace so you retain a vague sense of the wordshapes :) I could honestly see myself using this in a coffeeshop for the compromise between privacy and feedback.
125.
by efskap 1y ago
These multi modal LLMs are too expensive to run for that
126.
by efskap 1y ago
It worked for Amazon.com in 1995! > Bezos: We got an order from somebody in Bulgaria, and this person sent us cash through the mail to pay for their order. And they sent us two crisp $100 bills. And they put these two $100 bills inside a
127.
by efskap 1y ago
>needs location https://developer.mozilla.org/en-US/docs/Web/API/Geolocation...
128.
by efskap 1y ago
I tried to make the softphone approach work but I was unreachable far too often when Android decided to kill whichever softphone app I tried. And if it did keep running, I'm pretty sure it consumed decently more energy than a dedicated
129.
by efskap 1y ago
Yeah I dug up the relevant function because I was curious how one would solve this. return (r * 1000 + g * 100 + b * 10 + a * 255) / 10000; https://github.com/Stanko/monorail/blob/2da287a4e822705
130.
by efskap 1y ago
Cool stuff! Flattening things like colours into a 1D value is an interesting challenge here. How did you decide to weigh the R channel an order of magnitude higher than G and so on? I might've tried working with HSV for instance, since
131.
by efskap 1y ago
It's a dream of mine to learn enough stoneworking to turn a pair of stones from a local river into a mortar and pestle, ideally without contracting silicosis at the same time. Much cooler than shipping granite mined and cut on the othe
132.
by efskap 1y ago
Note that this also applies to GPT models on the API > That risk extended to users of ChatGPT Free, Plus, and Pro, as well as users of OpenAI’s application programming interface (API), OpenAI said. This seems very bad for their business.
133.
by efskap 2y ago
"Duolingo is Butter for Learning" because it prioritizes gamification and engagement over actual learning outcomes. Being joyful to use doesn't equate to effectiveness.
134.
by efskap 2y ago
The specific term for the shimmering mechanism is https://en.wikipedia.org/wiki/Binocular_rivalry
135.
by efskap 2y ago
The checkbox trick for the clickable Review button is so cool! Loved the easter egg links to various videos too
136.
by efskap 2y ago
Yes but this person learned to mimic the internet at large. Theoretically its preferred way of talking would be the average of all training data, as mimicry is GPT's training objective, and would therefore have very similar word distri
137.
by efskap 2y ago
Isn't this the attention mechanism, the reason we're using transformers for these things? Maybe not greater resolution per se, but focusing on a region with greater neural connectivity
138.
by efskap 3y ago
Did you even click the link? The "Why?" section explains the problem with tailwind and the code sample shows it's doing a different thing, closer to htmx.
139.
by efskap 4y ago
Sure it's a similar architecture but c'mon, 307 million learnable params for ViT-L vs 175 billion for GPT-3...
140.
by efskap 5y ago
> gaslighting That error page isn't making anyone question their perception or reality. And I'm pretty sure your average user associates browser error pages with problems "on their end", not site-specific ones. The pa
141.
by efskap 5y ago
> What's "king" if you subtract [male] without simultaneously adding [female]? Why does the model even consider that a possibility? I'd expect a gender-neutral term for a ruler, because "king" can be thought
142.
by efskap 5y ago
There's room for ambiguity. python egg, python length, python container, python types, python size, python lifespan
143.
by efskap 6y ago
Appearance. I'm just talking about typefaces - absolutely nothing to do with pronunciation. Some fonts do not have Cyrillic glyphs. If they could be generated by an adequately trained model based on the other glyphs, then that font&#x2
144.
by efskap 6y ago
Personally I'm curious about training a model to predict Cyrillic or Greek letterforms from Latin ones to boost multilingual coverage. I'm sure there are some learnable relationships there like Я <=> R or И <=> H.
145.
by efskap 6y ago
As an alternative approach, I mounted my 2012 Kindle PW, a raspberry pi, and my split keyboard (diverge 2) on a folding guitar stand. And I mean, it works: https://i.imgur.com/FNbKA0b.jpg Vim certainly makes the latency tol
146.
by efskap 8y ago
When it comes to Russian->English translation, the technology is most definitely not here.
147.
by efskap 8y ago
On the surface that's what it is, yes. It's a dark pattern with dynamic content though, and the linked articles explain it better than I can.
148.
by efskap 8y ago
Think this is the only other alternative for cross plat guis if you don't want to pull in non-.net stuff like qt or electron. https://github.com/AvaloniaUI/Avalonia So yeah.
149.
by efskap 9y ago
This doesn't require crouton or anything like that. https://skycocker.github.io/chromebrew/
150.
by efskap 9y ago
First they came for our Pepes, but I did not speak out, for I was not a frogposter.

More ›