• mpasila@alien.topB
    link
    fedilink
    English
    arrow-up
    1
    ·
    1 year ago

    Did anyone manage to get them working? I tried GGUF/GPTQ and running then unquantized with trust-remote-code and they just produced garbage. (I did try removing BOS tokens and still same thing)

    • Jelegend@alien.topB
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 year ago

      Yeah, exactly the same thing. Produced absolutely rubbish whatever i tried. I tried 8B 15B and 23B

    • watkykjynaaier@alien.topB
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 year ago

      I’ve completely fixed gibberish output on Yi-based and other models by setting the RoPE Frequency Scale to a number less than one, which seems to be the default. I have no idea why that works, but it does.

      What I find even more strange is the models often keep working after setting the frequency scale back to 1.