CUDA Mode Keynote | Andrej Karpathy | Eureka Labs

Поделиться
HTML-код
  • Опубликовано: 22 дек 2024

Комментарии • 18

  • @saturdaysequalsyouth
    @saturdaysequalsyouth 2 месяца назад +10

    This is the same talk that was posted on Latent Space, except the audio quality is much better

  • @iony_mikler
    @iony_mikler 2 месяца назад +1

    Your simulation seems to run very smooth. What is the computer you are using?

  • @dj0ai
    @dj0ai 2 месяца назад +7

    1:24 OMG! Andrej Karpathy lives in 2X. Me? I live in 0,75X

  • @ultrasound1459
    @ultrasound1459 2 месяца назад +1

    3:00 the goat 🐐 🙌

  • @eriye92
    @eriye92 2 месяца назад

    The GOAT

  • @sizwemamba689
    @sizwemamba689 2 месяца назад

    Excellent

  • @elliptictree
    @elliptictree 2 месяца назад +2

    Interesting.

  • @kehindeakeredolu3525
    @kehindeakeredolu3525 2 месяца назад

    first comment

  • @IvanMalison
    @IvanMalison 2 месяца назад +1

    i dont get why he's wasting his time doing this. Abstractions are good! Seems like a really strange thing to do.

    • @marshallmcluhan33
      @marshallmcluhan33 2 месяца назад +3

      Wasting time is what abstraction is all about. Slow tokens

    • @philippleser4125
      @philippleser4125 2 месяца назад +1

      No. Abstraction is not good per se. Most developers are brain washed that way. One of the reasons why Software has become bloated and sliw.Performance sucks. AK did the right thing.

    • @FamilyYoutubeTV-x6d
      @FamilyYoutubeTV-x6d 2 месяца назад

      I think you have a very valid point, sir. If we want to have "AI computers", we need to have performant and fast model training, execution, inference; otherwise, they are a waste of resources and not worth the hype or money. That's not to say that some "abstraction" is not good. ML and LLMs are not traditional software (in the deterministic sense), and given their goal of wanting to either mimic or emulate intelligence, some abstraction is bound to be necessary. Abstraction is crucial in mathematics and physics, as well as science in general.

    • @ultrasound1459
      @ultrasound1459 2 месяца назад +1

      Abstraction = Slow Ai

    • @FamilyYoutubeTV-x6d
      @FamilyYoutubeTV-x6d 2 месяца назад

      @@ultrasound1459 We are intelligent and have slow areas of intelligence too. Slow AI is not the bottleneck. Planning (and reasoning) is (are) one (two) of them.

  • @dontwannabefound
    @dontwannabefound 2 месяца назад +17

    I’ll save you 30 mins. He basically rewrites pytorch code to train gpt2 in c in 3000 LOC. there.