News for hackers. Stuff that matters.

A comment on Sub-1-Bit LLM Compression via Latent Factorization. Parent comment

Re:Sub-1-Bit LLM Compression via Latent Factorization by gobdovan on Thursday October 08, 2026 @03:09PM (#50010530)
> These projects exist, but aren't getting amplified.

Name them

[ Parent | on HN ]
Re:Sub-1-Bit LLM Compression via Latent Factorization by tempoponet on Thursday October 08, 2026 @04:48PM (#50012051)
Most popular would be Qwen 9b or Qwen 35b with experts offloaded to CPU. Even these aren't great, but they can be run at higher quants on lower hardware. We'll see what Qwen releases version 4 in the coming weeks.
[ Parent | on HN ]
Re:Sub-1-Bit LLM Compression via Latent Factorization by majorchord on Friday October 09, 2026 @09:57AM (#50020643)
there is also https://huggingface.co/prism-ml/Ternary-Bonsai-2-27B-gguf
[ Parent | on HN ]