Q2 GGUF versions?
#1
by
AKDesign
- opened
Can you upload Q2 GGUF @6GB versions of these models PLEASE ?!?!?!?!? or maybe even imatrix 2-4gb versions ???
At least Q2 = in size to SDXL, that most of us can run!
@17gb these models are FAT PIGS + 2gb L clip + 10gb LLM = 29gb thats huge memory hogs, most of our potato laptops can't handle it :(
Also, what make these models "uncensored" is it training on naughty content or is it that it uses "uncensored" LLM as text encoder?
If it's "uncesored" LLM that makes these models uncesored, does it mean using Un-LLM can be used with any other censored model to make it "uncensored" as well ie can I use Un-LLM on Q2 version I already have to make it "uncensored"???