you are viewing a single comment's thread
view the rest of the comments
[–] [S] 2 points 1 year ago (2 children)

32B, ikke 32 MB. Altså 32 milliarder parametre. Med 32-bit per parametre er det 128 GB.

  • source
  • parent
  • hideshow 4 child comments
  • [–] 3 points 1 year ago (1 child)

    OLMo 2 32B is the first fully-open model to outperform GPT3.5-Turbo and GPT-4o mini on a suite of popular, multi-skill academic benchmarks.

    Så dens ydeevne er vel ~1 år bagud ift. de proprietære state-of-the-art modeller. Ikke ringe for en komplet åben model.

    Kilde: https://allenai.org/blog/olmo2-32B

  • source
  • parent
  • hideshow 2 child comments