• David GerardOPMA
    link
    English
    193 months ago

    Basically there isn’t significant improvement to be had in the tokeniser, because it’s already been trained on all the data on earth. So all they have left is overengineering.

    • @UnseriousAcademic
      link
      English
      143 months ago

      Does this mean they’re not going to bother training a whole new model again? I was looking forward to seeing AI Mad Cow Disease after it consumed an Internet’s worth of AI generated content.

      • David GerardOPMA
        link
        English
        93 months ago

        I think they will do whatever gets more investor cash

      • @anton@lemmy.blahaj.zone
        link
        fedilink
        English
        83 months ago

        If you change the tokenizer you have to retrain from scratch, but you can do so with the old, unpolluted data.

        It’s genius if you think about it,* you can waste energy and tell your investors it’s a new better model, while staying upstream from the river you pollute.
        * at least for consultants, compute providers and other middle men.

        • @UnseriousAcademic
          link
          English
          43 months ago

          I remember one time in a research project I switched out the tokeniser to see what impact it might have on my output. Spent about a day re-running and the difference was minimal. I imagine it’s wholly the same thing.

          *Disclaimer: I don’t actually imagine it is wholly the same thing.

          • David GerardOPMA
            link
            English
            43 months ago

            there’s a research result that the precise tokeniser makes bugger all difference, it’s almost entirely the data you put in

            because LLMs are lossy compression for text

            • @froztbyte
              link
              English
              33 months ago

              latent space go brrrr

    • @corbin
      link
      English
      93 months ago

      Calling it now: codepoint-level non-tokenizing, with a remapping step to only recognize the most popular thousands of codepoints, would outperform what OpenAI has forced themselves into using. Evidence is circumstantial but strong, e.g. how arithmetic isn’t learned right because BPE tokenizers obscure Arabic digits. They can’t backpedal on this without breaking some of their API and re-pretraining a model, and they make a big deal about how expensive GPT pretraining is, so they’re stuck in their local minimum.

        • @UnseriousAcademic
          link
          English
          43 months ago

          The only viable use case, in my opinion, is to utilise its strong abilities in SolidGoldMagicarp to actualise our goals in the SolidGoldMagicarp sector and achieve increased margins on SolidGoldMagicarp.