Kimi Linear: An Expressive, Efficient Attention Architecture

 (github.com)

131 points | by blackcat201 8 hours ago

6 comments

  • eXpl0it3r 9 minutes ago
    For the uninitiated, what's a "hybrid linear attention architecture"?
  • amoskvin 5 minutes ago
    any hardware recommendations? how much memory do we need to this?
  • textembedding 34 minutes ago
    125 upvotes with 2 comments is kinda sus
    [-]
    • muragekibicho 2 minutes ago
      Lots of model releases are like this. We can only upvote. We can't run the model on our personal computers. We can neither test their 'Efficient Attention' concept on our personal computers.

      Honestly, it would take 24 hours just to download the 98 GB model if I wanted to try it out.

  • adt 5 hours ago
    https://lifearchitect.ai/models-table/
  • nekofneko 3 hours ago
    [flagged]
  • Ethan312 4 hours ago
    [flagged]