Thread

1. adroitboss · 1789503846 · focus · HN ↗

   I am positive I know exactly how this works, I made something similar a few months back. But the problem is without generation you are extremely limited in the use cases. And while the model can't hallucinate, it can still be wrong. It just can't make up data.

   1. dennisy · 1789503991 · focus · HN ↗

      Are you able to share how it works in that case?

      1. adroitboss · 1789506438 · focus · HN ↗

         I'll tell you this. Output isn't too cheap to meter, there is no decoder.

         1. krackers · 1789506838 · focus · HN ↗

            So an encoder-only model with a classifier trained on the heads or something?

            1. adroitboss · 1789555857 · focus · HN ↗

               This is basically what they have. <a href="https:&#x2F;&#x2F;github.com&#x2F;fastino-ai&#x2F;GLiNER2" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;fastino-ai&#x2F;GLiNER2

Unofficial Hacker News client; not affiliated with Y Combinator.