walrus01

18 ms·

481.
by walrus01 2mo ago
We should start asking it for an Albatross instead.
482.
by walrus01 2mo ago
It is interesting that it went with something like: "pelican = ocean bird = nautical theme = this means I should put a captain's hat on the cartoon Pelican" I think in some sense it understood the absurdity of a pelican ridin
483.
by walrus01 2mo ago
That is pretty good! It's interesting it chose orange, because that's actually a somewhat rare color for pelican cases. I would say 95% of the ones going around the world right now are either black or FDE (flat dark earth) tan co
484.
by walrus01 2mo ago
I think you may find that the dense 27B also does better if challenged with more rare coding tasks, less common or weird languages or things that aren't well represented in the active 3B parameters of the MoE model (eg: NOT css, javasc
485.
by walrus01 2mo ago
Try manually asking both more discrete esoteric knowledge questions. Or use benchmarks which are less coding focused. The 3.6-35B-A3B with post-training may do well in coding type benchmarks and math but the density of its knowledge falls o
486.
by walrus01 2mo ago
What's the result if you ask it for a "pelican equipment case" ? I've been trying the anti-bicycle pelican on some LLMs and the results are much more varied than the bicycle prompt. Some have very different ideas of pel
487.
by walrus01 2mo ago
And then somewhat later, when things moved to 100Mbps, the 3c905b and 3c905c-tx
488.
by walrus01 2mo ago
Yeah, I see what you mean now, I am looking at it on mobile and that animated slider to show or hide the entire menu seems like it takes unnecessarily long.
489.
by walrus01 2mo ago
Some of it looks like tailwind css and they have a syntax highlighting addon that shows up in the homepage source as well.
490.
by walrus01 2mo ago
I haven't looked at it on mobile, that's unfortunate to hear. I don't see any animation when interacting with it on a desktop browser, snapping open a left side header menu item to see sub-items seems instant.
491.
by walrus01 2mo ago
Sort of on a tangent to the actual content of the page, but what documentation system/CMS is this particular web page UI? It seems very fast and relatively uncluttered for the menu on the left side, content pane, top horizontal menu ba
492.
by walrus01 2mo ago
> Answering 2,500 independent questions is an embarrassingly parallel workload, all it needs is scale out. I believe the assumption of the benchmark creators is that it's meant to measure the sequential speed that a single instance
493.
by walrus01 2mo ago
well, yeah, it's based on a 2+ year old tiny model. It's very much an alpha proof of concept that they can perma-bake an LLM into silicon. https://huggingface.co/meta-llama/Llama-3.1-8B
494.
by walrus01 2mo ago
The news article is really not clear about whether this was, relative to the company "OSS": a) OSS is a colocation customer with its own hardware colocating it inside an Iron Mountain datacenter. In which OSS owned the bare metal
495.
by walrus01 2mo ago
1.7T
496.
by walrus01 2mo ago
I am fairly sure the blog post (and everything else on that site) was written by an LLM and the larger ones with deep training data sets do know what a lombard loan is, but it was instructed to write the content to match a certain aesthetic
497.
by walrus01 2mo ago
I don't mean anything I wouldn't expect to see, just that it's a very well known thing, and also has had some recent challenges (news articles about extreme climate, etc). And similar stories going back 12+ years such as: ht
498.
by walrus01 2mo ago
Googling "agricultural products as collateral cheese" turns up some very interesting results. https://www.google.com/search?q=agricultural+products+as+col...
499.
by walrus01 2mo ago
Erik Prince and similar have been trying to get that going again (for actual commercial cargo shipping related activities) for some time now.
500.
by walrus01 2mo ago
I wonder if instead of blocking the facebook ads, in a continual game of whack-a-mole with facebook changing things, there would be a way to accept the ads and then generate a white opaque box overlay where the ad is rendered in the page.
501.
by walrus01 2mo ago
Mass automated vulnerability scans have been a very common thing since years before the advent of this in 2001: https://en.wikipedia.org/wiki/Code_Red_(computer_worm) I remember when 'code red' spread and it
502.
by walrus01 2mo ago
Just FYI lm-studio is a GUI wrapper on top of a copy of llama-server that the lm-studio developers compile and distribute
503.
by walrus01 2mo ago
Please give it a try before you assume it won't work? Tell me what platform you're on and follow the three sequential cli commands I provided, and if it doesn't compile I'll be glad to look at your error message and get
504.
by walrus01 2mo ago
This is like complaining that the mariadb team has failed to make their tech easy to use. It's a server side piece of software. There are plenty of user friendly GUIs or wrappers for it. lm-studio or unsloth are to llama.cpp as somethi
505.
by walrus01 2mo ago
I only want linkedin cringe if it's of the artisanal, hand crafted, organic, free range, GMO free, fair trade, heirloom variety.
506.
by walrus01 2mo ago
It's literally three steps, assuming you have the equivalent of the debian "build-essential" dependencies installed on your system for cmake, compiler. It's the exact three rows of my post, pasted one at a time into the
507.
by walrus01 2mo ago
At this point the options are llama-server or vLLM if you're serious about running things at your desk in the under 256GB RAM size class (70B, 120B size models). In addition to, of course, 27B to 35B size things. With of course a ton o
508.
by walrus01 2mo ago
I mean it's probably not, I just haven't got used to it yet. It's about the same level of security as installing a windows app on win2000 25 years ago and blindly downloading a .exe off the internet and running it to get into
509.
by walrus01 2mo ago
One way I prevent possible catastrophic fuckups is that the 'doing code work' box that runs opencode or pi or whatever, is its entirely own separate VM and desktop environment (running as a xen or kvm guest and with its own LVM lo
510.
by walrus01 2mo ago
the full set of llama.cpp binaries builds in under 5 minutes with an unmodified build workflow straight from their github page on a literally ten year old dual xeon.

More ›