‹ BackHN Continuity

Thread

Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms

575 points · 225 comments · firelex

  1. k__ · · focus · HN ↗
    Von 1.2 had a better Doom score :D

    <a href="https:&#x2F;&#x2F;github.com&#x2F;wfzyx&#x2F;von" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;wfzyx&#x2F;von

    1. nico · · focus · HN ↗
      Oh wow, those Doom scores for Jeff are pretty terrible

      The Von numbers have led me on a rabbit whole of getting a classifier to play Doom

      I got it to average 22 kills (the max is 26) on that same scenario that Jeff and Von are testing on (it’s called Defend Center)

      Now I’m having it play a more advanced scenario, and it’s doing about 45 kills (SOTA is ~59 kills)

      It’s amazing what you can do with small classifiers if you can collect some data. These models I’m testing train on CPU in seconds (what takes the longest is running the game, doing test runs and collecting data), they are &lt;1MB in size and do inference in &lt;1ms on CPU

      Edit: after looking at Jeff&#x27;s numbers more in detail, the 6.5 kills number is not that bad, but it can definitely be better ;)

      1. kridsdale1 · · focus · HN ↗
        One day we’ll be using the same Kills&#x2F;SOTA metric for models driving physical kill-bots.
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.