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/)

31337

@ 31337 @sh.itjust.works

Posts
7
Comments
73
Joined
3 yr. ago

  • how much power does your system need?

    Jump

  • The PC I'm using as a little NAS usually draws around 75 watt. My jellyfin and general home server draws about 50 watt while idle but can jump up to 150 watt. Most of the components are very old. I know I could get the power usage down significantly by using newer components, but not sure if the electricity use outweighs the cost of sending them to the landfill and creating demand for more newer components to be manufactured.

  • The Trek Place

    Jump

  • A lot of people seem to use "DEI" as a stand-in for the "n-word."

  • Watching the Generative AI Hype Bubble Deflate

    Jump

  • I'm loading up on vacuum tubes.

  • Deleted

    Permanently Deleted

    Jump

  • Last time I looked it up and calculated it, these large models are trained on something like only 7x the tokens as the number of parameters they have. If you thought of it like compression, a 1:7 ratio for lossless text compression is perfectly possible.

    I think the models can still output a lot of stuff verbatim if you try to get them to, you just hit the guardrails they put in place. Seems to work fine for public domain stuff. E.g. "Give me the first 50 lines from Romeo and Juliette." (albeit with a TOS warning, lol). "Give me the first few paragraphs of Dune." seems to hit a guardrail, or maybe just forced through reinforcement learning.

    A preprint paper was released recently that detailed how to get around RL by controlling the first few tokens of a model's output, showing the "unsafe" data is still in there.

  • Elon Musk suddenly realizes he has no clue how to govern

    Jump

  • He can use his money to support challengers in their next elections. He's already said he's going to do this to the Democrats. As we've seen with AIPAC recently, it doesn't take much money to unseat politicians. Musk has enough money to basically control congress.

  • Deleted

    Permanently Deleted

    Jump

  • I think TikTok appeased the right by changing their algorithm. Charlie Kirk is apparently doing extremely well on the platform now.

  • Serving my media library to my TV (local network only), i need suggestions

    Jump

  • You can also just install the libreelec os (Kodi), and install the Jellyfin Kodi addon. Haven't tried that addon. I used to use Kodi when I had only one TV, and liked it. Now that I have 2 Android TVs, just installing Jellyfin on the TVs works fine. I might go back to rPIs and disconnect my TVs from the internet though.

  • Totally happened

    Jump

  • It's not really automation though. The store is outsourcing labor to the consumer.

  • Stop using generative AI as a search engine

    Jump

  • I use GPT (4o, premium) a lot, and yes, I still sometimes experience source hallucinations. It also will sometimes hallucinate incorrect things not in the source. I get better results when I tell it not to browse. The large context of processing web pages seems to hurt its "performance." I would never trust gen AI for a recipe. I usually just use Kagi to search for recipes and have it set to promote results from recipe sites I like.

  • OpenAI, Google, Anthropic admit they can’t scale up their chatbots any further

    Jump

  • Hmm. I just assumed 14B was distilled from 72B, because that's what I thought llama was doing, and that would just make sense. On further research it's not clear if llama did the traditional teacher method or just trained the smaller models on synthetic data generated from a large model. I suppose training smaller models on a larger amount of data generated by larger models is similar though. It does seem like Qwen was also trained on synthetic data, because it sometimes thinks it's Claude, lol.

    Thanks for the tip on Medius. Just tried it out, and it does seem better than Qwen 14B.

  • OpenAI, Google, Anthropic admit they can’t scale up their chatbots any further

    Jump

  • Larger models train faster (need less compute), for reasons not fully understood. These large models can then be used as teachers to train smaller models more efficiently. I've used Qwen 14B (14 billion parameters, quantized to 6-bit integers), and it's not too much worse than these very large models.

    Lately, I've been thinking of LLMs as lossy text/idea compression with content-addressable memory. And 10.5GB is pretty good compression for all the "knowledge" they seem to retain.

  • Hyperspace: a p2p VPN solution that doesn't require a server

    Jump

  • I haven't checked it out in years. From my understanding, IPFS aims to be a distributed filesystem that kinda works like Bittorent. If you access a file, you then seed it. Last time I checked it out, the project was jumping on the crypto bandwagon... Just checked out their website now, and don't know WTF it is.

  • Bernie Sanders doubles down that people are ‘angry’ with Dems after Pelosi said she didn’t ‘respect’ his remarks

    Jump

  • That's not really good evidence. They could've made more money by just holding an SP 500 index fund. Their recent Visa trade is better evidence.

  • Calculatable

    Jump

  • NCalc+ is pretty good. I don't use it for anything complicated though.

  • Teary-eyed John Oliver begs reluctant voters to back Harris

    Jump

  • Another interesting question is if they vote for one of their local DA or sheriff candidates or just obtain.

  • A 14-year-old took home $10,000 for his award-winning investigation into train derailments. Here's what he found.

    Jump

  • Meh, startups and businesses are capitalist organizations, and I think the idea of patents is questionable outside capitalism, so these wouldn't really be a good metrics. I'd guess the richest countries "innovate" the most because they can support more risky endeavors. The U.S. is the capitalist imperial core, so it probably innovates the most. Other capitalist nations like Haiti, probably not so much.

    The best measure of innovation would probably be something like scientific publications. China wins by raw numbers, Vatican City wins per-capita (???).

  • Spider Staredown

    Jump

  • Yeah, I mostly leave Harvestmen alone if they stay out of the way. I kill brown recluses though, because I've heard their bites can do some damage.

  • Trump says RFK Jr. plan to remove fluoride from public water 'sounds okay to me'

    Jump

  • It’s not alarmist: A second Trump term really is an extinction-level threat to democracy

    Jump

  • tankie

    🤨 I've never heard a right-winger, or even a liberal use that term.