batperson

7 ms·

1.
by batperson 9d ago
The future of inference is likely in ASICs, so we'll get the inverse, a bit less capable than frontier but super fast models. Like this 14k tok/s beast https://chatjimmy.ai/ from Taalas (who got acquired by AMD re
2.
by batperson 12d ago
Do you know what reasoning level this was generated at?
3.
by batperson 15d ago
The video compression created the illusion, it removed enough frames to where you can't tell the motion of the spokes since every frame refresh they moved enough distance to be "valid" for both forward and backward motion.
4.
by batperson 1mo ago
I had 3x32" 4k at 100% scaling for the longest time (32" 4k is still the best bang for buck right now imo). Main horizontal center display and a vertical on each side, angled aggressively towards my head. But 4k is quite weird for
5.
by batperson 2mo ago
Buy what? Hardware for inference? Pretty sure you'd still be giving money to "AI assholes", just somewhat different flavor. And you'd likely have to pay so much that the $60/mo would seem like peanuts, and in the en
6.
by batperson 2mo ago
I own an elgato Stream Deck (somewhere in a drawer), I love the concept of keys being a display but the keys are VERY mushy. Still a better deal and a way more versatile device than that Codex Micro pad. Now that I think about it, I think I
7.
by batperson 2mo ago
What about people with disabilities, physical or mental, who can't get out, old people with no family or family who doesn't care? In theory it would be great if everyone got some attention and socialized, in reality a lot of peopl
8.
by batperson 2mo ago
Social people will be fine, I think this tech is far more important for lonely people who for any reason don't get to socialize much (if at all), this is especially common in older people. These people might not have any other alternat
9.
by batperson 3mo ago
Edge user here. For one, chromium is faster than firefox, any given page will load about 20% faster, another reason is edge workspaces feature, I've grown to like it, which seems to be some sort of chromium feature that everyone bakes
10.
by batperson 4mo ago
If you check openrouter there are a tons of providers selling API access to open source LLMs at a fraction of the cost compared to SOTA models (codex/claude). What model you're serving and what kind of platform you serve is a big
11.
by batperson 4mo ago
Before they removed it, I was using groq Kimi K2 model for a chat bot in small community site/chat. It was really good, seemed to have incredibly vast general world knowledge and the fast speed (400tok/s if I remember right) meant
12.
by batperson 7mo ago
These are business class laptops, there's no dedicated GPU. Where are you're going to utilize this high refresh rate? I'm pretty sure 99% of the time the integrated graphics would be working hard to churn out 120 frames of st
13.
by batperson 1y ago
Damn, what a flashback. I forgot about that game, it was quite something for its time. I remember the gesture spell casting system not working very consistently, but it was still a ton of fun.
14.
by batperson 2y ago
I found the opposite when I went through a split keyboard phase and did Colemak-DH layout. As far as english goes I found that a lot of words had a smoother "flow" when it comes to finger location. In the end I went back to a regu
15.
by batperson 2y ago
I can't find information on what would happen to Chromium, would google need to hand it over too? That's probably more important than the fate of Chrome.
16.
by batperson 2y ago
+1, a nicer looking ff experience, more customizable too.
17.
by batperson 2y ago
Wasn't aware of "Claude Dev", just tried it out, it's pretty cool. I had this simple node chatbot with a handful of features but it was all just one big script, asked it to split up the logic into separate files that wou
18.
by batperson 2y ago
Nuts, seeds and olive oil are all extremely high in calories.
19.
by batperson 2y ago
That's the old/regular model. This post is about the new "instruct" model.
20.
by batperson 3y ago
A big issue that these articles never mention is that using browser dev tools to debug CSS becomes a huge pain with tailwind. It also makes it harder for community to create custom themes/user scripts for your website/app since ta
21.
by batperson 3y ago
Then you can think of Spotify as a more user friendly alternative to Napster, Kazaa, Soulseek, Limewire and friends.
22.
by batperson 3y ago
The PPI ratio for 32" 4k with no scaling is actually pretty good (at least to my eye, since it's similar to my previous 24" 1080p, 27" 1440p monitors). If we can agree on that, then the solution is to just move the monit
23.
by batperson 3y ago
Sidebery is great for this nowadays, looks pretty slick too. https://addons.mozilla.org/en-US/firefox/addon/sidebery/
24.
by batperson 3y ago
Scrobbles... reminds me of this: https://www.youtube.com/watch?v=ZR8az1UMOiQ
25.
by batperson 3y ago
Could you elaborate ? Does this require any specific setup?
26.
by batperson 3y ago
You can also just click the . (period) key when viewing any repo.
27.
by batperson 3y ago
Here's my custom dark mode CSS if anyone enjoys the google dark mode aesthetic (I inject with the "Stylus" extension, avoid "Stylish" extension at all costs): https://pastebin.com/CsVkRbYM Also my C
28.
by batperson 4y ago
In the case of Washington Post it just has a "position: fixed" style on the <body> element. That's usually the case with most of these scroll locking sites, one of the root parent elements will have some CSS style that
29.
by batperson 4y ago
When it comes up '@apply' I don't really it as an improvement even over vanilla CSS. Seems like tt that point you're basically learning an entirely new syntax just to save time on typing a few characters.
30.
by batperson 4y ago
This sounds like a ditch digger trembling in fear upon seeing an excavator.

More ›