Live data from Hacker News

Viewing profile — frontierkodiak

frontierkodiak

HN member
Joined
Mon, Jan 17, 2022, 6:21 PM UTC
HN karma
65
Public activity
36 items

About frontierkodiak

Working to make ecology more effective by engineering away its bottlenecks @ Polli Labs.

clbswrs.co (caleb at) polli.ai

Recent public activity

  1. comment
    Comment #48691514

    Maybe not 10x; but it's fantastically talented at intuiting intent, reconstructing contexts, and making aligned judgement calls. You could throw Fable utter garbage and get great r…

  2. comment
    Comment #48323638

    At 6x the cost of its predecessor!

  3. comment
    Comment #47669100

    Apologies; I do not mean to imply any affiliation.

  4. comment
    Comment #47632600

    See my other comment in this thread! I've got you covered; launching open models in ~1month. Happy to answer any questions!

  5. comment
    Comment #47632574

    Hey, good news! Pollination ecologist + ML guy here; with open models coming soon. You can keep an eye on (gh) polli-labs/linnaeus (a bit stale; I'll rebase on my private repo late…

  6. comment
    Comment #45161392

    Rip2 is another useful alternative: https://github.com/MilesCranmer/rip2 I have a small wrapper around rip2, aliased to `recycle`; files go to a `graveyard` zfs dataset. I deny `rm…

  7. comment
    Comment #43537744

    oh man that's awesome. I have been working for quite some time on big taxonomy/classification models for field research, espec for my old research area (pollination stuff). the #1 …

  8. comment
    Comment #43021309

    I've owned 17 Seagate ST12000NM001G (12TB SATA) drives over the last 24mos in a big raidz3 pool. My personal stats, grouping by the first 3-4 SN characters: - 5/8 ZLW2s failed - 1/…

  9. comment
    Comment #41001402

    Sure. The talk about 8bit refers to quantization-aware training. Pretty common in image models these days to reduce the impact of quantization on accuracy. Typically this might mea…

  10. comment
    Comment #40360381

    I haven't managed to successfully export my custom ViT model yet, but I've not had an issue accessing the export methods in torch 2.3 within the nvcr.io/nvidia/pytorch:24.02-py3 co…

  11. comment
    Comment #39046728

    I've pasted QR codes on the backs of bees before! Unfortunately we started this project a week before COVID so we didn't get to see it through, but the actual gluing is easier than…

  12. comment
    Comment #38975528

    YMMV, but I have tried this several times and have found it very difficult without microsoldering equipment. Some ESP variants have a built-in U.FL connector which makes things eas…

  13. comment
    Comment #37861429

    Not to mention that you can trivially set a 300w power cap on the 4090 and still get 75-80% of peak FLOPS.

  14. comment
    Comment #37188559

    Tropical Tidbits is excellent.

  15. comment
    Comment #36368474

    Various hibernation modes are great, but 300ma@5v sometimes feels too hungry for the amount of computation in there. I bet things would be better if it wasn’t 45nm ‘pitch’. They ar…

  16. comment
    Comment #35878102

    He was a technical writer at Boeing, IIRC

  17. comment
    Comment #35211716

    We don’t have the data to make much more than directional claims re: macro trends in insect biodiversity. Ecological monitoring is tedious and labor-intensive; insects are simply t…

  18. comment
    Comment #34851264

    Firefox reliably remembers my tabs, but never my pinned tabs-- which, of course, are the ones that I actually want to be preserved across restarts. In my dream world, pinned tabs p…

  19. comment
    Comment #34645881

    FF with Sidebery is nearly perfect, except that it seems that persisting pinned tabs across sessions is an alien technology that nobody has quite figured out.

  20. comment
    Comment #34306358

    The things others are pointing out (esp. selective grazing) are all factors. Not sure how well-evidenced this is, but something I've heard from relevant researchers is that bison h…

  21. comment
    Comment #33940138

    The article states “The first base edit disabled the T-cells targeting mechanism so they would not assault Alyssa's body”. Not an expert but I presume the ‘targeting mechanism’ is …

  22. comment
    Comment #32932519

    Incredible biodiversity monitoring— everywhere, all the time

  23. comment
    Comment #31953379

    That's the main approach I've encountered. Slice, inference, filter duplicates from the results. https://github.com/obss/sahi

  24. comment
    Comment #31773616

    My friend, shoot me an email and I’d love to chat. My present situation is not all that dissimilar to what you describe. I have been in a MSc program for the past 2 years which has…

  25. comment
    Comment #31735804

    It is silly that we naively define diversity by what is trivially measurable, which tends you can determine of a person without knowing anything of their person. Give me a campus w…