woadwarrior01

5 ms·

31.
by woadwarrior01 1mo ago
> SSDs as Memory They've implemented mmap() in hardware. What a time to be alive?! nb: Related discussion on Samsung's PIM from yesterday: https://news.ycombinator.com/item?id=49487341
32.
by woadwarrior01 1mo ago
Because OP is running it on an M3 Ultra with Ollama. He'll get much better perf with llama.cpp or MLX, both of which Ollama wraps, albeit very poorly.
33.
by woadwarrior01 1mo ago
It's a software issue.
34.
by woadwarrior01 1mo ago
I suspect this might be due to MTP mispedictions. Either because the 3.8 quantized model weights do not include MTP heads, or as what happened with Ornith-1.5 recently, corrupted MTP heads, or due to a software issue in Ollama.
35.
by woadwarrior01 1mo ago
This would've likely saved Steve Jobs, had it arrived 15 years earlier.
36.
by woadwarrior01 1mo ago
Captive audience.
37.
by woadwarrior01 1mo ago
That benchmark is super sus. Until someone pointed it out, the top performing open weights model was a Kimi K3 fine tune from their sponsor (abacusai/Smaug-Agentic). Now, it's not on the list. Source: https://twitterweb
38.
by woadwarrior01 1mo ago
It's widespread here in Ireland. POS terminals have 5%, 10%, 15% and 20% tip options with a tiny and hard to reach dismiss button.
39.
by woadwarrior01 1mo ago
FWIW, macOS has good sandboxing, but LMStudio, Ollama, Darkbloom etc aren't sandboxed. This is also the reason why none of these things aren't distributed via the Mac App Store, because the Mac App Store mandates sandboxing.
40.
by woadwarrior01 1mo ago
I find it mildly ironic that the author's name means Hope in Persian.
41.
by woadwarrior01 2mo ago
RTN quantization of weights
42.
by woadwarrior01 2mo ago
Check out Clean Links[1]. It does everything Hyperduck does and more (supports multiple Macs) for free. Disclaimer: I'm the author of the app. [1]: https://cleanlinks.app/
43.
by woadwarrior01 2mo ago
OT: The 2nd link is the first time I've seen a Microsoft FrontPage website in decades. :)
44.
by woadwarrior01 2mo ago
I'm an app developer and all my apps have the "Data Not Collected" privacy nutrition label. I love the idea, but Apple's enforcement of it is very lackadaisical. I've reported dozens of apps that were blatantly lyi
45.
by woadwarrior01 2mo ago
I have a co-worker who symlinks CLAUDE.md to AGENTS.md in all the repositories he owns.
46.
by woadwarrior01 2mo ago
There's an unspecified short/random delay between the user taking the screenshot and the notification firing. The trick used by Bluesky, Telegram etc is using a different technique. They (mis)use secure input fields to show an ove
47.
by woadwarrior01 2mo ago
Really happy to see this. Kudos to Qualcomm for releasing it under a liberal Apache 2.0 license. Strategically speaking, I think this only makes sense as an anti-NVIDIA play.
48.
by woadwarrior01 2mo ago
If you're referring to GDPR, companies routinely evade such protections using "informed consent" / "legitimate interests" loopholes. The big ones get caught once in a while, get a slap on the wrist and continue
49.
by woadwarrior01 2mo ago
Yes, iOS has userDidTakeScreenshotNotification for that. https://developer.apple.com/documentation/uikit/uiapplicatio...
50.
by woadwarrior01 2mo ago
I'm only peripherally familiar with Go. IIRC, Rust defaults to move-only structs with the option to make them copyable with the Copy trait. Swift defaults to copyable structs (like Go), but has a ~Copyable generic type constraint to ma
51.
by woadwarrior01 2mo ago
Most languages would encode such behavior in a trait or a protocol instead of a zero-length struct field. It's type information and ought to be encoded in the type. From that perspective, I think it is a hack.
52.
by woadwarrior01 2mo ago
> noCopy is a special marker for types that must not be copied after their first use. if it looks like a hack, walks like a hack, and quacks like a hack...
53.
by woadwarrior01 2mo ago
NVIDIA's 3090 was released in September 2020. Apple's M1 was released ~2 months later. Anyone with a 5 year old M1 Mac with 32GB or more RAM can run a 4-bit quantized version of Qwen3.8 27B on their machine. AFAICT, there are ~110
54.
by woadwarrior01 2mo ago
I agree 100%. IMO, it all started with ollama misrepresenting the Deepseek R1 distills as Deepseek R1, all for hype and marketing. I've had so many ostensibly technical people telling me: "I tried DeepSeek R1 and it was terrible&q
55.
by woadwarrior01 2mo ago
I think writing is the killer use case for local LLMs. We've had so many advancements in LLM samplers for improved text generation (off the top of my head: min-P, adaptive-P, XTC, DRY, p-less, Top-H, Top-n-Sigma, and so many more) but
56.
by woadwarrior01 2mo ago
> The distillations of Qwen 3.8 down to a 27B model are good, but they're not on a par with frontier models. Does it have to be? There are plenty of coding tasks, where it's good enough.
57.
by woadwarrior01 2mo ago
I think his point is hand wavy at best. It presupposes infinite scaling and ignores all the algorithmic efficiency wins that are being discovered. Ironically, many of which are being discovered with autoresearch style workflows, using the v
58.
by woadwarrior01 2mo ago
Because subtraction is not a total operation on positive integers. Negative numbers leave the domain.
59.
by woadwarrior01 2mo ago
Perhaps some of the people who got into traditional leadership roles got there due to factors besides their raw leadership abilities. LLMs are an equalizer, and with them, the inability to lead them simply becomes a "skill issue",
60.
by woadwarrior01 2mo ago
There are some very interesting latent compaction approaches like this[1] for when you can control the whole inference stack. i.e in on-device and datacenter inference. [1]: https://arxiv.org/abs/2602.16284

More ›