‹ BackHN Continuity

Thread

Show HN: Maple-Preview – Ternary 20B MoE running at 120 tok/s on a iPhone

169 points · 52 comments · edwardbzhang

  1. getcrunk · · focus · HN ↗
    Anyone compare this to ternary bonsai vs the 1 bit
    1. woadwarrior01 · · focus · HN ↗
      I haven't compared it for perf, but at first glance this appears to be an LLM (text-only) while both binary and ternary bonsai 27B models are VLMs (where the vision tower and the adapter MLP weights are unquantized and kept in float16).
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.