‹ BackHN Continuity

Thread

I built non-autoregressive decision models with RL a year ago

1363 points · 319 comments · nandakishor_ml

  1. tarruda · · focus · HN ↗
    At this size (~400 million parameters), does it become viable running directly on CPU?
    1. rcarmo · · focus · HN ↗
      Yep. Not instantly though. I am hacking away at these things over on <a href="https:&#x2F;&#x2F;github.com&#x2F;rcarmo&#x2F;go-pherence" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;rcarmo&#x2F;go-pherence (I do SIMD versions of common inference algos) and trying to improve that.
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.