llama-fit-params: lower ctx size for multi GPU#18101

JohannesGaessler merged 1 commit intoggml-org:masterggml-org/llama.cpp:masterfrom

JohannesGaessler:llama-fp-fix-multi-gpu-ctxJohannesGaessler/llama.cpp:llama-fp-fix-multi-gpu-ctxCopy head branch name to clipboard