I compared the costs between luna and haiku for some classification work. Luna comes out a fair amount cheaper due to more efficient tokenization and less outputs. Model input tokens output tokens batch cost sync cost…
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"
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 online,…
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 risk that…
> 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…
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…
(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…
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's…
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 times on…
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…
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…
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'd…
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…
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…
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…
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…
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…
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"…
...could be named for how much space is left after that soft core has been instantiated.
Out of interest, what could make such a design better than a traditional inductive motor? Such motors already suffer inductive losses, and do not need slip rings or brushes, and presumably do not an additional set of…
I know press releases never describe anything that's actually novel, but it looks like they're just describing an asynchronous induction motor? I'm wondering if this release is implicitly pretending that some "low slip…
Im assuming the "halt and catch fire" thing is only really possible on pre-microprocessor machines built of discrete components, where the individual parts are small enough to overheat by themselves if driven wrong. I'd…
just a few drive-by comments: First and most important, the title claims these are safety-critical rules, but the conclusion says that they are only being used on mission-critical projects. I get the feeling that there…
Honestly I think your observation is mostly right... if you don't have a hard performance requirement, then rust probably isn't the best language to write your code in...
GP might have been referring to undefined/invalid behaviour (whether in the language or in some OS syscall or whatever). After the demons came out of your nose you can never fix the problem, so there is no point trying…
I compared the costs between luna and haiku for some classification work. Luna comes out a fair amount cheaper due to more efficient tokenization and less outputs. Model input tokens output tokens batch cost sync cost…
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"
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 online,…
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 risk that…
> 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…
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…
(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…
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's…
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 times on…
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…
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…
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'd…
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…
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…
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…
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…
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…
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"…
...could be named for how much space is left after that soft core has been instantiated.
Out of interest, what could make such a design better than a traditional inductive motor? Such motors already suffer inductive losses, and do not need slip rings or brushes, and presumably do not an additional set of…
I know press releases never describe anything that's actually novel, but it looks like they're just describing an asynchronous induction motor? I'm wondering if this release is implicitly pretending that some "low slip…
Im assuming the "halt and catch fire" thing is only really possible on pre-microprocessor machines built of discrete components, where the individual parts are small enough to overheat by themselves if driven wrong. I'd…
just a few drive-by comments: First and most important, the title claims these are safety-critical rules, but the conclusion says that they are only being used on mission-critical projects. I get the feeling that there…
Honestly I think your observation is mostly right... if you don't have a hard performance requirement, then rust probably isn't the best language to write your code in...
GP might have been referring to undefined/invalid behaviour (whether in the language or in some OS syscall or whatever). After the demons came out of your nose you can never fix the problem, so there is no point trying…