Run on a cluster?

[bonze82]

Is it possible to run gpt4all on a linux based cluster to make gpt4 faster, I have 6 systems that run a openMPI cluster and would like to be able to run gpt4all across the cluster with 24gigs on the main system and 8 gigs on the other 5 systems.

[cebtenzzre]

GPT4All is based on llama.cpp, which has OpenMPI and OpenSHMEM support. You should try their main example with one of those backends to see how it works for you. There is active development here:
ggml-org/llama.cpp#3334
ggml-org/llama.cpp#4571