☆ Yσɠƚԋσʂ ☆@lemmygrad.ml to technology@hexbear.netEnglish · 3 days agoGPT-5, the world best model just 1 year ago, is today inferior to Qwen3.6 27B that you can run on your desktoplemmygrad.mlimagemessage-square29fedilinkarrow-up176arrow-down10
arrow-up176arrow-down1imageGPT-5, the world best model just 1 year ago, is today inferior to Qwen3.6 27B that you can run on your desktoplemmygrad.ml☆ Yσɠƚԋσʂ ☆@lemmygrad.ml to technology@hexbear.netEnglish · 3 days agomessage-square29fedilink
minus-squarelurkerlady [she/her]@hexbear.netlinkfedilinkEnglisharrow-up6·edit-23 days agousually the merge models, uncensored ones, abliterated, etc are complete dogshit. if you can, run unsloth ggufs on llama.cpp and anything of vital importance in transformers
usually the merge models, uncensored ones, abliterated, etc are complete dogshit. if you can, run unsloth ggufs on llama.cpp and anything of vital importance in transformers