SirDimples@programming.dev to LocalLLaMA@sh.itjust.worksEnglish · 1 month agoGLM-5.3 Imminent dropgithub.comexternal-linkmessage-square10linkfedilinkarrow-up10arrow-down10file-text
arrow-up10arrow-down1external-linkGLM-5.3 Imminent dropgithub.comSirDimples@programming.dev to LocalLLaMA@sh.itjust.worksEnglish · 1 month agomessage-square10linkfedilinkfile-text
minus-squareleanleft@lemmy.mllinkfedilinkEnglisharrow-up0·1 month agoon the extreme opposite side of the spectrum: https://huggingface.co/LiquidAI/LFM2.5-2.6B or… really extreme… https://huggingface.co/AxiomicLabs/GPT-X2.5-135M
minus-squareJrockwar@feddit.uklinkfedilinkEnglisharrow-up0·29 days agoAre these any good? I mean, within their size class, obviously - not expecting them to compare to Kimi K3! Models the size of that 135M one open up some interesting use cases for edge devices.
minus-squarehumanspiral@lemmy.calinkfedilinkEnglisharrow-up0·21 days agoit’s by far best sub trillion parameter model. fits in 512gb with 1m context at q4. Or less rental hardware than any model that scores higher than it.
quite large though!
on the extreme opposite side of the spectrum:
https://huggingface.co/LiquidAI/LFM2.5-2.6B
or… really extreme… https://huggingface.co/AxiomicLabs/GPT-X2.5-135M
Are these any good? I mean, within their size class, obviously - not expecting them to compare to Kimi K3!
Models the size of that 135M one open up some interesting use cases for edge devices.
it’s by far best sub trillion parameter model. fits in 512gb with 1m context at q4. Or less rental hardware than any model that scores higher than it.