Thread

1. abraxas · 1789680452 · focus · HN ↗

   I'm not following the local mdoel scene too closely but this seems quite amazing. Is this able to be run on Apple silicon too?

   1. kamranjon · 1789680506 · focus · HN ↗

      "Ternary Bonsai 2 27B reaches up to 143 tokens/second on NVIDIA GeForce RTX 5090 and 46.8 tokens/second on M5 Max. On an RTX 4090, Ternary Bonsai 2 27B consumes just 0.714 mWh/token, making it 40% more energy-efficient than an 8B model running in full-precision."

      1. pizza234 · 1789681997 · focus · HN ↗

         Their mention of the 5090 is bit odd, since on 32 GB GPUs, Q6 fits while having better quality. Very interesting model for 16 GB GPUs though!

         1. sisve · 1789683442 · focus · HN ↗

            They mention 5090 with regards to speed, Q6 will not have that speed?

            And speed matters a lot for many use cases

            1. selectodude · 1789685848 · focus · HN ↗

               150 tokens per second on a ternary model implies that it’s GPU bound, I’d bet a Q6 model is even faster because it’s existed longer and seen more optimization. You’d have to be insane to not run an NVFP4 quant over a ternary quant on Blackwell if they both fit.

Unofficial Hacker News client; not affiliated with Y Combinator.