zhouwul
Results
2
comments of
zhouwul
Thank you! I notice in llama.cpp line 15904 shows "// TODO: needs fix after https://github.com/ggerganov/llama.cpp/pull/3228 " . It seems like GGML_USE_MPI codes haven't updated. Maybe MPI still has been broken....
Now i successfully run MPI 4.1.2 on branch b1100(I try it just because it's an old branch)