News for hackers. Stuff that matters.
Re:Unsloth Qwen3.8-27B Quantizations – Dynamic v3.1 Update by xrd on Saturday November 14, 2026 @09:25AM (#50497254)
I've been happily using unsloth/Qwen3.6 for a while using llama.cpp on several rtx3090s. Excited to try this update.

And, reading through the ggml update on the front page of HN this morning, I realize I might do better than llama.cpp.

I've also really fascinated by llama-modchip and running these models in RAM (https://github.com/synchopate/llama-modchip.git)

Is there a place to find the best way to optimize my hardware setup?

I know about this:

https://github.com/noonghunna/club-3090

Are there others? Are there better nooks to look?

I assume unsloth studio (since it uses llama.cpp behind the scenes) is not going to do better than HEAD of llama.cpp from github, but am I wrong?

Re:Unsloth Qwen3.8-27B Quantizations – Dynamic v3.1 Update by strix_varius on Saturday November 14, 2026 @11:14AM (#50498182)
> And, reading through the ggml update on the front page of HN this morning, I realize I might do better than llama.cpp.

What do you mean by this? (from another llama.cpp user)

Re:Unsloth Qwen3.8-27B Quantizations – Dynamic v3.1 Update by xrd on Saturday November 14, 2026 @11:34AM (#50498345)
Re:Unsloth Qwen3.8-27B Quantizations – Dynamic v3.1 Update by danielhanchen on Sunday November 15, 2026 @01:16AM (#50504459)
We're innovating as well so there definitely will be more updates! We have a fork where we also add edits for eg https://github.com/unslothai/llama.cpp/releases/tag/b12347-m... has a lot of perf fixes and also other model arch support which mainline does not have