boutell

10 ms·

91.
by boutell 8mo ago
The speed is ridiunkulous. No doubt. The quantization looks pretty severe, which could make the comparison chart misleading. But I tried a trick question suggested by Claude and got nearly identical results in regular ollama and with the ch
92.
by boutell 8mo ago
Strong meme game. I'm on an older release and now I'm reluctant to update. In my current release, the verbosity is just where I want it and control-o is there when I really need it.
93.
by boutell 8mo ago
Flawless logic! I have a spoiler-tastic fan theory about the movie Marty Supreme that is apropos here.
94.
by boutell 8mo ago
This stuff is cool. So is whisper. But I keep hoping for something that can run close to real time on a Raspberry Pi Zero 2 with a reasonable English vocabulary. Right now everything is either archaic or requires too much RAM. CPU isn'
95.
by boutell 8mo ago
Thank you. The waste heat problem is so bad but no one gets around to mentioning the fact that you can't have AI grade chips and space at the same time.
96.
by boutell 8mo ago
That's a solution for minuscule amounts of heat that nevertheless disturb extremely sensitive scientific experiments. Using gold, no less. This does not scale to a crapton of GPU waste heat.
97.
by boutell 8mo ago
How do you cool them? Getting rid of heat is one of the number one challenges on the ISS.
98.
by boutell 9mo ago
The most important benchmark: https://boutell.dev/misc/qwen3-max-pelican.svg I used Simon Willison's usual prompt. It thought for over 2 minutes (free account). The commentary was even more glowing than the image.
99.
by boutell 9mo ago
I feel like the time for this was two years ago, and LLMs are now less bothered by remembering syntax than I am. It's a nice lisp-y syntax though.
100.
Try out Apple's ML-sharp 2d photo to 3D converter on your own computer(github.com)
1 points by boutell 9mo ago|1 comments
101.
by boutell 9mo ago
For those who would like to try it for real and not wait in a queue. It's a very fast model, even with your CPU it doesn't take very long. This repo is a convenience wrapper that sets up Python just so, runs ml-sharp, and then bui
102.
by boutell 9mo ago
Knowing all of these is exactly what a developer shouldn't need to do. Fix "big O" problems in your own code. And be aware of a few exceptionally weird counterintuitive things if it matters on a "big O" level — like
103.
Easily create and view 3D splat files from 2D images with Apple's ML Sharp model(github.com)
4 points by boutell 10mo ago|0 comments
104.
by boutell 10mo ago
I like Claude but this is an absolutely tone deaf thing on Anthropic's part. I've been pondering that given what the inputs are, llms should really be public domain. I don't necessarily mean legally, I know about transformati
105.
by boutell 11mo ago
I love programming in listicle: HERE ARE THE TOP 10 fibbonacci_numbers: YOU WON'T BELIEVE n := 6
106.
by boutell 11mo ago
He lost me a bit at the end talking about running chat bots on CPUs. I know it's possible, but it's inherently parallel computing isn't it? Would that ever really make sense? I expected to hear something more like low end con
107.
by boutell 11mo ago
Google owns 14% of Anthropic and Anthropic is using Google TPUs, as well as AWS Trainium and of course GPUs. It isn't necessary for one company to create both the winning hardware and the winning software to be part of the solution. In
108.
by boutell 11mo ago
Had me going for a minute there.
109.
by boutell 11mo ago
This is absolutely the first thing I looked for too. They just barely mentioned thermal management at all. Maybe they know something I don't, but I know from past posts here that many people share this concern. Very strange that they d
110.
by boutell 1y ago
Not sure why so much love for the idea the article is slop. I suffer through slop regularly and this article didn't press that button for me at all.
111.
by boutell 1y ago
Don't miss out on zooming in, each one is visualized so you can tell the non Starlink ones apart.
112.
by boutell 1y ago
Shut off the broken bot filter so we can read it please
113.
Big trouble at Linode's Newark data center(status.linode.com)
3 points by boutell 1y ago|2 comments
114.
by boutell 1y ago
While most of our newer customers are on AWS, we have a number of legacy customers on Linode/Akamai. Sunday morning, our VPSes started going offline. At 6:08am Eastern on Sunday, Linode's status page acknowledged big trouble &quot
115.
by boutell 1y ago
Point of information: I believe this project uses Tauri, which actually does use web technology and even JavaScript for rendering, it just does it with the native web renderer of the platform so you're not dragging around a superfluous
116.
by boutell 1y ago
To be fair, they wanted it very cold indeed
117.
by boutell 1y ago
An old idea, impractical not maybe not absolutely impossible. Somehow they made it worse
118.
by boutell 1y ago
This is very, very cool. Impressive work. I'm interested to see whether the final feature set will be larger than what you'd get by creating a type-safe language with a pythonic syntax and compiling that to native, rather than bui
119.
by boutell 1y ago
I see what you did there! There's a LISP Machine with its guts on display at the MIT Museum. I recall we had one in the graduate student comp sci lab at University of Delaware (I was a tolerated undergrad). By then LISP was faster on a
120.
by boutell 1y ago
Huh. This was my entry point as well, but I find Linux very aligned with my feelings about the 4.1 experience, as opposed to Solaris which felt clunky and overbuilt. Of course this is purely subjective emotional stuff I'm talking about

More ›