10 comments

  • ilusion 9 minutes ago
    Have you tested what it remembers from early in the stream after a shift in the topics thrown at it?
  • whizzter 1 hour ago
    Nobody will throw rocks, I think most people are curious/suspicious about the big players and wants more hands-on since we suspect that this all will come down in cost soon enough.
  • advael 59 minutes ago
    Seems interesting, I've been messing with a lot of continuous learning approaches lately and it's cool to see something that's built from the ground up for avoiding catastrophic forgetting. Worth a clone for sure
  • cpldcpu 37 minutes ago
    Is this architecture actually able to generalize or is it mostly based on memorization? Have you tried some basic tasks that require generalization? e.g. number addition etc?
    • volotat 24 minutes ago
      The model is way too small and undertrained to make any generalization claims. I want to wait until it reads the whole corpus I gave and then test it on some simple established benchmarks to see how it will behave.
      • jacquesm 8 minutes ago
        What kind of hardware are you using for training?

        nm, I found it:

        > RTX 3070 Laptop GPU with 8 GB

        Super impressive.

  • skeledrew 1 hour ago
    Getting conceptually closer to how the human brain works. Looking forward to more of this.
    • volotat 1 hour ago
      I also like how it is very organic. It naturally grows and deletes unused elements, so in addition to traditional backprop there is also a natural selection happening in the background. Each new expert has 16 parents by the way, lol.
  • hexley19 1 hour ago
    Seeing 'Mini-AGI' and '8GB VRAM' in the same sentence is a breath of fresh air. Maybe local AGI isn't so far-fetched.
  • hanselot 31 minutes ago
    THANK YOU SO MUCH. This is the missing piece.
  • loopydosuette 42 minutes ago
    throwing crumpled paper ball
  • myshapeprotocol 1 hour ago
    [dead]