ONLY gguf models supported.

// https://raw.githubusercontent.com/ggerganov/whisper.cpp/master/ggml/ggml_vk_generate_shaders.py

To start from cli:
cd /opt/GPT4All/bin
./chat 

or from cli:
LD_LIBRARY_PATH=/opt/GPT4All/bin/ /opt/GPT4All/bin/chat

or from menu;office (XFCE,PLASMA,GNOME,etc)


Its a full GPT4All installation, not only chat, lot of tips exist in /usr/local/bin/GPT4All speciall for python guys...
You are free to maintain it in the SBo way and simply remove my name, no copyrights

Deps are in slack-required file

If you upgrade from older version to new after package build finish just do the old time classic 
# upgradepkg /tmp/gpt4all-2.6.1-x86_64-1_rtz.txz

This way your old models and localldocs will stay in place ;) 


Just a few notes to add from me Condor.

OpenBLAS is an optimized BLAS (Basic Linear Algebra Subprograms) library and if you will play with AI it's good to have it.
For PPL with AMD video card could use my repo at https://www.ixip.net/rocm/ROCm-6.1/x86_64/ to download and install
AMD ROCm Library for Slackware64-current. I'm trying for now to refresh packages on every minor version, because the 
compilation proccess require a lot's of time and server which I don't have. 

If you are on desktop machine with AMD and don't want to install ROCm, I recommend you to compile the project with VULKAN support.

I don't have NVIDIA video card so I can't help there.