Show HN: We compressed a 31B video+audio diffusion model to 14B, 6 steps, INT8(github.com)
1 pointby jiant39 minutes ago1 comment
  • jiant39 minutes ago
    [flagged]