Show HN: Run 30B model in 4GB Active Memory

 (github.com)

4 points | by vkkhare 1 day ago

1 comments

  • nrjpoddar 1 day ago
    Link github/sparse_transformers seems to be broken
    [-]