Live data from Hacker News

Viewing profile — declaredapple

declaredapple

HN member
Joined
Fri, Dec 22, 2023, 7:11 PM UTC
HN karma
287
Public activity
143 items

About declaredapple

No profile information was provided.

Recent public activity

  1. comment
    Comment #39934146

    Some people may be willing to travel but don't know where the best place to go is.

  2. comment
  3. comment
    Comment #39896738

    It's going to be GPT-3.5turbo not GPT4. As others have mentioned, binggpt/bard are also free. These smaller models are a relatively cheap to run, especially in high batches. I'm su…

  4. comment
    Comment #39866974

    They've been designing their own chips a while now, including with an NPU. Also because of their unified memory design, they actually have insane bandwidth which is incredibly usef…

  5. comment
    Comment #39866942

    The flash doesn't do the computations though, that's just a method of getting it to the processor

  6. comment
    Comment #39856753

    I think there's clear social impairment. Making friends is harder and it takes more "manual" effort to socialize "effectively". However much of this social impairment may or may no…

  7. comment
    Comment #39856650

    > We're currently working with professional to improve our domain rating, ensuring you receive the quality backlinks needed to boost your SEO. Please be patient. It might take a wh…

  8. comment
    Comment #39853377

    > companies like OpenAI have had access to large quantities of H100 for a few months now and Sora is being presented From what I could tell from Nvidia's recent presentation, Nvidi…

  9. comment
    Comment #39844449

    What? Are you asking if the framework automatically quantizes/prunes the model on the fly? Or are you suggesting the LLM itself should realize it's too big to run, and prune/quanti…

  10. comment
    Comment #39842027

    For me GPT4 seems to suggest more generic unittests in python. They're much more "Put tests here", or "path.to.dependency". Claude3 opus (and often Sonnet) actually fills in the fu…

  11. comment
  12. comment
    Comment #39841226

    > you do not divulge details of your training data. FWIW asking LLMs about their training data is generally HEAVILY prone to inaccurate responses. They aren't generally told exactl…

  13. comment
    Comment #39794283

    > But the biggest reason why PostgreSQL would not change its license is the disservice it would do to all PostgreSQL users. It takes a long time to build trust in a technology that…

  14. comment
  15. comment
    Comment #39783972

    > NASA, which follows the federal government’s General Schedule pay scales, offers starting salaries along a range that starts at $54,557 for engineers with bachelor’s degrees, $66…

  16. comment
    Comment #39782513

    Heads up that tickets generally don't guarantee the plane you'll be on, they can change at any time for a bunch of different reasons.

  17. comment
    Comment #39759511

    AMD desperately needs to improve their software situation. It's been like this for 10 years - they have competitive hardware in many cases - but the software is always a buggy mess…

  18. comment
    Comment #39758938

    I assume it's similar to the 1 billion row challenge with java - C# and Java have sufficiently advanced VM's that they can compete with C/CPP performance - assuming you go out of y…

  19. comment
    Comment #39747990

    I think that happens because talc deposits often contain asbestos so it's not an uncommon contaminate. that doesn't mean talc products always contain asbestos https://www.ncbi.nlm.…

  20. comment
    Comment #39719039

    I'm not completely disagreeing. But it is one of the few "engineering fields" where many of the top performers don't have degrees or formal educations. It would be a shame to gatek…

  21. comment
    Comment #39717962

    Is this a question or a statement? All ChatGPT does is predict the most likely next token (with a little sampling on the top most likely tokens). That's it. It doesn't lookup wheth…

  22. comment
    Comment #39684894

    Dark sky used to be accurate almost to the minute for me. Apple Weather will tell me it won't rain today or all week. Meanwhile NOAA will tell me I'm currently in a thunderstorm an…

  23. comment
    Comment #39684809

    What the actual crap did Apple do to mess it up so bad then? Switching between providers on Carrot, Apple Weather often doesn't predict any amount of rain for the entire week, mean…

  24. comment
    Comment #39671088

    Take # parameters and divide by 2 and then add like 1 to estimate it for 4bit quantization. A 13B model will need ~7GB of memory at 4bit. Most quantization models start falling apa…

  25. comment
    Comment #39671044

    In this case it appears to be for SEO spam to gain backlinks