Most active commenters

    ←back to thread

    Llama.cpp 30B runs with only 6GB of RAM now

    (github.com)
    1311 points msoad | 11 comments | 31 Mar 23 20:37 UTC | HN request time: 1.217s | source | bottom
    Show context
    detrites ◴[31 Mar 23 20:58 UTC] No.35393558[source]▶
    The pace of collaborative OSS development on these projects is amazing, but the rate of optimisations being achieved is almost unbelievable. What has everyone been doing wrong all these years cough sorry, I mean to say weeks?

    Ok I answered my own question.

    replies(5): >>35393627 #>>35393885 #>>35393921 #>>35394786 #>>35397029 #
    1. politician ◴[31 Mar 23 21:25 UTC] No.35393885[source]▶
    Roughly: OpenAIs don’t employ enough jarts.

    In other words, the groups of folks working on training models don’t necessarily have access to the sort of optimization engineers that are working in other areas.

    When all of this leaked into the open, it caused a lot of people knowledgeable in different areas to put their own expertise to the task. Some of those efforts (mmap) pay off spectacularly. Expect industry to copy the best of these improvements.

    replies(5): >>35393957 #>>35394249 #>>35394957 #>>35396490 #>>35397974 #
    2. bee_rider ◴[31 Mar 23 21:32 UTC] No.35393957[source]▶
    The professional optimizes well enough to get management off their back, the hobbyist can be irrationally good.
    replies(1): >>35395341 #
    3. hedgehog ◴[31 Mar 23 22:00 UTC] No.35394249[source]▶
    They have very good people but those people have other priorities.
    4. dsjoerg ◴[31 Mar 23 23:04 UTC] No.35394957[source]▶
    ~Whats a jart?~

    Ah I see https://news.ycombinator.com/user?id=jart

    replies(1): >>35395087 #
    5. sp332 ◴[31 Mar 23 23:18 UTC] No.35395087[source]▶
    In March 2014, Tunney petitioned the US government on We the People to hold a referendum asking for support to retire all government employees with full pensions, transfer administrative authority to the technology industry, and appoint the executive chairman of Google Eric Schmidt as CEO of America

    https://en.m.wikipedia.org/wiki/Justine_Tunney

    replies(1): >>35396194 #
    6. fallous ◴[31 Mar 23 23:47 UTC] No.35395341[source]▶
    The professional operates within prioritization tranches. Make it work, make it reliable, make it fast, make it pretty. If you're still iterating on proof-of-concept/prototyping you'll generally confine yourself to the first and/or second levels. Once you've settled on a finalized prototype you then follow the rest of the prioritization levels to achieve shippable product.
    7. hackernewds ◴[01 Apr 23 01:44 UTC] No.35396194{3}[source]▶
    what's your point? also interestingly JART is in the thread here, so they might have read your comment :)
    replies(1): >>35401864 #
    8. dflock ◴[01 Apr 23 02:39 UTC] No.35396490[source]▶
    The jart: the standard unit of developer.

    "I've got 20 yrs experience, and I think I'm about 150 milli jarts, maybe 200 on a good day."

    9. fy20 ◴[01 Apr 23 07:13 UTC] No.35397974[source]▶
    I'd say it's probably not a priority for them right now.

    Of course it would save them some money if they could run their models on cheaper hardware, but they've raised $11B so I don't think that's much of a concern right now. Better to spend the efforts on pushing the model forward, which some of these optimisations may make harder.

    replies(1): >>35403525 #
    10. sp332 ◴[01 Apr 23 16:59 UTC] No.35401864{4}[source]▶
    I'm pretty sure she knows already?
    11. VernorVintage ◴[01 Apr 23 19:55 UTC] No.35403525[source]▶
    It's a pretty big concern if you had to spend a billion on training, but 6 months later the open source community is able to replicate your training for <100K because you were too cheap to hire an extra 100 optimization experts

    That'd be a 10,000 fold depreciation of an asset due to a preventable oversight. Ouchies.