Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)S

SuspiciousCarrot78

@ SuspiciousCarrot78 @aussie.zone

Posts
18
Comments
318
Joined
3 mo. ago

  • Self hosted photo options for the family?

    Jump

  • Are you optimizing for a polished experience or for simplicity?

    If for something polished that works very well, you've probably already found the best solution - immich. There's also Piwigo, Photoprism etc.

    If you're looking for something simpler, can I suggest using syncthing (think: like a local version of Dropbox, sorta), with PiGallery2?

    https://en.wikipedia.org/wiki/Syncthing

    https://github.com/bpatrik/pigallery2

    The idea here is that it creates a website that you can access on your Wi-Fi that is populated by the files that you push into the folder using your sync thing.

    That's not as elegant as immich, but it's another option if you're looking for low-end hardware.

  • Homelab Documentation

    Jump

  • Reliable messenger for family use?

    Jump

  • Molly still have SMS integration? It use to have ability to act as an SMS/MMS client.

  • OpenAI admits responsibility for HuggingFace Attack - an agent from an internal evaluation is reportedly the cause

    Jump

  • Ha! I was just coming to post this.

    Yeah, this is fucked.

    It's the equivalent of taking someone's hand, punching them in the face with it and telling them to stop hitting themselves.

    Worse than that, this is a cyber attack by a frontier, close source lab on the bastion of open source AI.

    Whether that's incompetence or maliciousness or just a PR stunt, I don't know, but it stinks to high heaven.

    As Louis Rossman has recently become fond of saying "govern yourself accordingly".

  • Apple Plans to Defeat RAM Prices by Letting You Lease a Mac

    Jump

  • It's nice when the bad guy wears a black hat and twiddles their mustache, isn't it? Really helps to speed things up.

  • Homelab Documentation

    Jump

  • I just break things enough times until I internalize the knowledge into my soul.

    Kidding aside, I do wish I was more disciplined about documentation.

    It could have really helped me earlier in the week in fact. I figured out something really clever a couple of months ago, promptly wiped the computer. And now I can't recreate it. I've come up with a halfway solution, but it's not as elegant as it used to be.

    To be honest with you, this is a really good use case for a large language model - self hosted or other wise. You know Karpathy's LLM-wiki method?

  • Drop OSS - an open-source, self-hosted alternative to platforms like Steam and Epic.

    Jump

  • ...and the fact that their website uses AI generated templates, colours, fonts, call to action sure makes me believe that uncritically :/

    I don't care to spelunk their code base, but the old "how you do one thing is how you do everything" adage comes to mind.

    EDIT: fwiw

    https://threadiverse.link/lemmy.ml/comment/26821243

  • Drop OSS - an open-source, self-hosted alternative to platforms like Steam and Epic.

    Jump

  • Chinese startup Moonshot AI unveils Kimi model it says rivals OpenAI, Anthropic

    Jump

  • I was looking thru an old box of bits yesterday for a project - I needed 2x8GB sticks of ddr3, bought circa 2023. Lo and behold, they had sent me 2x16 DDR4. Wrong size, wrong type, wrong timing.

    I'd be mad but apparantly I'm now the king of Londinium and can afford a shiny hat.

  • Chinese startup Moonshot AI unveils Kimi model it says rivals OpenAI, Anthropic

    Jump

  • It's one of those "if you have to ask, you can't afford it" scenarios.

    I've heard FP16 is likely to be 6TB of VRAM, when Kvcache is factored in.

    Even a more sensible quant will likely be 1TB.

    ... most of us are not running this at home.

  • I gave my local LLM agent a face — an avatar frontend with an "emotion-beat" output contract (open-source)

    Jump

  • Apple M7 Ultra Chip Planned With Up to 1.5 TB of Unified Memory

    Jump

  • Yes. People occasionally talk about pooling resources / creating a co-op to buy something like this. Easier to split $200K purchase (and probably $10k/month electricity costs) if 200-2000 people chip in. But then...that's just the cloud with extra steps.

    Can't says I've ever seen a co-op like that work but ICBW

  • Apple M7 Ultra Chip Planned With Up to 1.5 TB of Unified Memory

    Jump

  • Yeah. But they're slow because most of us are GPU peasants. If someone were willing to drop $3-5K on a rig, they could probably run decent, dense models at greater than cloud speeds. Hell, with enough black magic, they could do it with less, but they'd have to go deep into the weeds.

    OTOH, $3-5K buys you a shit ton on Open Router, Claude, Chat, Lumo etc.

    The game is entirely rigged for "you will own nothing and be happy about it".

  • Apple M7 Ultra Chip Planned With Up to 1.5 TB of Unified Memory

    Jump

  • I guess it depends on your definition of perfect - cheap, good or fast.

    This thing is probably going to cost at least $20K USD.

    Edit:

    "Next year's base M7 processor is expected to arrive in the first half of 2027 and will also upgrade memory bandwidth to about 240 GB/s."

    That's...really fucking slow. What's the goal here - CGI, engineering sims, game dev etc? 1.5TB is cool but at 240GB/s that will crawl for AI use.

    Comparison: this is about $100K, for 7.2TB/s, 252GB VRAM (+500GB system ram, so closer to 750GB total)

    https://www.nvidia.com/en-us/products/workstations/dgx-station/

  • AI giants learn what everyone else on the modern internet already knows

    Jump

  • Honestly, I think that's the only llm of its kind. The next closest thing is this

    https://github.com/karpathy/nanochat

    If you don't mind something narrower (an expert system, sort of), I'm slowly coding something for myself here. It's not ready yet (about half way done to initial release).

    https://codeberg.org/BobbyLLM/picoGURU

    Public facing docs are a bit of a mess at the moment (sorry; writing docs is boring AF; will probably be the last thing I tidy before official 1.0.0) and it's not really runnable right now, but when done, the "source files" for this will be kiwix (off line Wikipedia), your own markdown notes and 30 or so trusted domains (user defined).

    So it's all either your own content or public domain aka open source fully. Code itself is AGPL-3.

    Anyway, that's the goal.

  • AI giants learn what everyone else on the modern internet already knows

    Jump

  • Rayfish, Iroh and Yggdrasil

    Jump

  • It also says -

    "So it's all vibe-coded slop."

    The idea is over 4 years old: a trustless, decentralized network with no central entity that can censor who you connect with. I sketched it for years. What an LLM changed was the speed: the prototype came together in a day. The trigger was the release of iroh v1.

    Make of that what you will.

  • Keep Android Open (Stop Google from limiting APK file usage)

    Jump

  • Well sure, if you're installing CFW or aosp and you don't need the ever increasing cruft that seems to rely on Play store authentication. Never install GApps, done.

    Some folks (not you or I) need to access play store to install their apps (I think even WhatsApp and Signal need it now; I couldn't install it on my phone), 2FA, transport apps, bank apps etc.

    Catch 22 for them.

    • degoogled phone....but can't install all the apps they need
    • Google ecosystem... but at whims of Google.

  • Nomad: ESP-32 media server

    Jump

  • How will your phone stream movies to the TV, little billy's tablet and Grandma's laptop, while letting Suzie listen to her favourite music and Davey read his CBR comics?

    What does your phone draw, BTW, when watching media or casting it? Is it less than 0.8 of a watt?

    Can your phone fit in a matchbox?

    Ya dig? :)