cm0002@lemmy.world to LocalLLaMA@sh.itjust.worksEnglish · 6 days agoShould be able to load the full version of DeepSeek R1 on this no prob 😎😎lemmy.worldimagemessage-square9fedilinkarrow-up122arrow-down13
arrow-up119arrow-down1imageShould be able to load the full version of DeepSeek R1 on this no prob 😎😎lemmy.worldcm0002@lemmy.world to LocalLLaMA@sh.itjust.worksEnglish · 6 days agomessage-square9fedilink
minus-squareSmokeyDope@lemmy.worldMlinkfedilinkEnglisharrow-up1·edit-24 days agoYou are correct in your understanding. However the last part of your comment needs a big asterisk. Its important to consider quantization. The full f16 deepseek r1 gguf from unsloth requires 1.34tb of ram. Good luck getting the ram sticks and channels for that. The q4_km mid range quant is 404gb which would theoretically fit inside 512gb of ram with leftover room for context. 512gb of ram is still a lot, theoretical you could run a lower quant of r1 with 256gb of ram. Not super desirable but totally doable.
You are correct in your understanding. However the last part of your comment needs a big asterisk. Its important to consider quantization.
The full f16 deepseek r1 gguf from unsloth requires 1.34tb of ram. Good luck getting the ram sticks and channels for that.
The q4_km mid range quant is 404gb which would theoretically fit inside 512gb of ram with leftover room for context.
512gb of ram is still a lot, theoretical you could run a lower quant of r1 with 256gb of ram. Not super desirable but totally doable.