Llama.cpp 30B runs with only 6GB of RAM now

Great to see this advancing! I’m curious if anyone knows what the best repo is for running this stuff on an Nvidia GPU with 16GB vram. I ran the official repo with the leaked weights and the best I could run was the 7B parameter model. I’m curious if people have found ways to fit the larger models on such a system.

I'd assume that 33B model should fit with this(only repo that I know of that implements SparseGPT and GPTQ for LLaMa), I, personally, haven't tried though. But you can try your luck https://github.com/lachlansneff/sparsellama

looks great thank you!