Quant request
Could you make this one with i8mm quants, please? https://huggingface.co/ResplendentAI/Aura_Uncensored_l3_8B
It has the outdated L3 tokenizer, so existing quants for this model don't work.
Also this one, when you have free time, please: https://huggingface.co/Casual-Autopsy/L3-Luna-8B
Sorry about that, I must have missed your first message (hard to keep track when only logging in a couple times a day 😅)
Both are now in the queue :)
Sorry about that, I must have missed your first message (hard to keep track when only logging in a couple times a day 😅)
Both are now in the queue :)
Thank you man, you are a hero.
Could you make this one with i8mm quants, please? https://huggingface.co/ResplendentAI/Aura_Uncensored_l3_8B
It has the outdated L3 tokenizer, so existing quants for this model don't work.
It seems that Aura Uncensored didn't make it through the queue. Perhaps because it has the outdated original L3 tokenizer and it gave an error?
@bartowski
could you add this to the queue, please?:
https://huggingface.co/Undi95/Lumimaid-Magnum-12B
https://huggingface.co/bluuwhale/L3-SthenoMaid-8B-V1
Your i8mm's are a God-send for running on Layla.
Ah yeah it did have an error it seems.. that's super annoying, I may try it again is I can fix it but it may be a lost cause :(
I'll add those two :)
Ah yeah it did have an error it seems.. that's super annoying, I may try it again is I can fix it but it may be a lost cause :(
I'll add those two :)
Don't worry man. I don't know how to fix the old tokenizer problem either. The last problem quanting I saw was this, but I don't know if it's the same: https://huggingface.co/nbeerbower/llama-3-SNAMD-8B/discussions/1#6659df6d377e197825fd88cd
Thank you for adding those two!!