Farmadupe

4 ms·

1.
by Farmadupe 7d ago
ah, yes that's obviously fine; I'm definitely coming fromt he perspective of "bored guy on a Saturday evening who happened to stumble across someone else's wall of charts on HN"
2.
by Farmadupe 7d ago
I'd actually recommend some excellent books on the "philosophy" of data presentation: The first that comes to mind is "the visual display of quantitative information" by Edward Tufte seems to be freely available onl
3.
by Farmadupe 7d ago
Yes exactly. The issueThe epherrality of the VMs isn't an issue, it's the _shared_ part that's the concern here. Going by the fact that the kernel is listed as "kernel 7.0.0-1012-azure" I feel like it's a fair
4.
by Farmadupe 7d ago
> compare shapes and ratios, not absolute MB/s In this case, given that the author's own disclaimer (above) already disclaims the numeric readings, I'm not sure how it's possible to make any inference on "shapes
5.
by Farmadupe 7d ago
> CI runs use loop devices on shared ephemeral VMs (one VM per filesystem): compare shapes and ratios, not absolute MB/s. Each job records a host-calibration anchor — see the table. I think if you're not using baremetal for suc
6.
by Farmadupe 7d ago
> 2G of random garbage is written directly onto one member device (behind the filesystem's back, offset 1G — python injector; uutils dd mis-seeks on dm devices), caches dropped, then a full scrub: btrfs scrub -B, zpool scrub + wait,
7.
by Farmadupe 7d ago
@farlight assuming that you're the creator do you think you'd be able to rework the HTML/CSS? I'm sure you've got good data but speaking on behalf of my eyeballs, the results page is... hard to read!
8.
by Farmadupe 13d ago
Thankyou for putting the last three months of my life into words <3 > and the explanations of the agent are convincing enough that surely, it knows better than you. I feel this in my bones. I also get to watch the misalignment feedbac
9.
by Farmadupe 18d ago
(If it helps, I ask my own question of myself too -- I mostly don't write code by hand any more as I find that an LLM writes it faster and with less bugs -- Is that therefore proof that my time was never worth my paycheck? I hope not b
10.
by Farmadupe 18d ago
hmm, assuming that this article is part written by claude and part human-written, can anyone help me find a rule of thumb for "how to know if the article is worth reading"? Because on the one hand, the prose and the presentation i
11.
by Farmadupe 23d ago
Don't know why you're getting faded out, because having worked on exactly this engine, software mitigations were put in place. That's exactly your `anomaly detection system` It was a belt-and-braces fix, because the primary f
12.
by Farmadupe 24d ago
Thankyou! that title was painful, I spent 20 seconds term rewriting it in my head.. * Firefox's AI Switch Is Off. Telemetry Isn't * Firefox's AI Switch Is Off. Telemetry Isn't off [complete second clause] * Firefox'
13.
by Farmadupe 1mo ago
It's a valid shift to move onto actually trying to read the article critically (which I don't mean in an insulting way -- If you assume a writing has something worthwhile to tell you, reading it critically is how you learn the wor
14.
by Farmadupe 1mo ago
Yeah, I'm with you on this, I think this is just what fable/opus-5 slop looks like now... - "A frozen verify.py accepts the claim" (what does it mean to freeze a python script?) - "which trains the recipe eight time
15.
by Farmadupe 3mo ago
I'm totally feeling you here as well. Fable is definitely quicker to first output token on `claude.ai` (ie less internal reasoning tokens being generated), and given how much more expensive decode is than prefill, I'm sure that mu
16.
by Farmadupe 3mo ago
Yes this is an extremely well known result for exactly the reason you guessed. It's not just abcktracking, asking an LLM to present a conclusion and then justify is also an excellent way to provoke hallucination as the model con concts
17.
by Farmadupe 3mo ago
Is it too cynical to read the first quote as "when openai catch up, we will restore fable 5 to subscription plans"? In other words, they're saying "We know have a monopoloy and we will take all of your money until someo
18.
by Farmadupe 3mo ago
2 and 3 bit quants are often closer to gibberish than hallucination, and that'll happen regardless of context. I shouldn't claim too much, I haven't tried GLM5.2 at 2/3 bit quantization, but if I were a betting man I&#x2
19.
by Farmadupe 4mo ago
It's amazingly vacuous isn't it? I think the most interesting read was the fact that they were surprised llama.cpp crashed when they used a bad set of commandline arguments. Although in the section immediately above the observatio
20.
by Farmadupe 5mo ago
I'm not sure there's a one-stop shop for this at the moment. I think the process is: * Have a box with sufficient spare (V)RAM -- probably 8G for simple categorization with qwen3.5-4b, and 24G or more for more intelligent categori
21.
by Farmadupe 5mo ago
Yup, especially when for a lot of us, the price of the frontier subscription has become a cost of doing business over the last 6 months. If you're already doing big boy stuff with big boy models, then... just carry on trucking! Only pl
22.
by Farmadupe 5mo ago
The latest rounds of open weights vision language models are incredibly good. Like, massively good. Open weights vision capabilities trade blows with frontier models. Over the last few months I'd roughly rank capabilities as Gemini -&g
23.
by Farmadupe 5mo ago
If it helps, I mean it in a really literal sense. qwen3.6 27b is currently $3.20 per million tokens on openrouter right now which is way overpriced. As good as the 27b is, kimi k2.5 $3.00 and it's just in another league in terms of cap
24.
by Farmadupe 5mo ago
I wonder if for a model that small with a permissive license it might not be worth their time to host a commercial grade inference stack? Might be easier to chuck it over the fence and let other providers handle it as it'll run in almo
25.
by Farmadupe 1y ago
Using protobuf is practical enough in embedded. This person isn't the first and won't be the last. Way faster than JSON, way slower than C structs. However protobuf is ridiculously interchangeable and there are serializers for eve
26.
by Farmadupe 2y ago
I feel like there's a semi-philosophical question somehwere here. Recursion is clearly a core computer science concept (see: many university courses, or even the low level implementation of many data structures), but it's surprisi
27.
by Farmadupe 3y ago
The author is making a deliberate point about undefined behaviour in the article. Hence them not executing worked examples. In fact, by not doing so they are making a subtle implicit statement that it is uninteresting to consider actually
28.
by Farmadupe 3y ago
FWIW, 85k 4-input luts is huge by the standards of any softcore with "micro" in the name. It comfortably surpasses the capacity of most of the Actel aerospace FPGAs that I tend to work with. And I think it's so "micro&qu
29.
by Farmadupe 3y ago
...could be named for how much space is left after that soft core has been instantiated.
30.
by Farmadupe 3y ago
Steppers are super optimized for being super precise and easy to controle while being (relatively) cheap to build. They have weird "trick" geometry that means you can move the rotor to many different angles (tens, hundreds) with f

More ›