gpt4all-cuda: fix CUDA initialization

[haras-unicorn]

Description of changes

Fixes #321928.

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandboxing enabled in nix.conf? (See Nix manual)
  • sandbox = relaxed
  • sandbox = true
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 24.11 Release Notes (or backporting 23.11 and 24.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
• Fits CONTRIBUTING.md.

---

Add a 👍 reaction to pull requests you find important.

[Titaniumtown]

I do not have an nvidia gpu to test with. So I do not have any way of testing this.

[Titaniumtown]

@VeilSilence Since you reported the original issue. Do you have a sec to test this? Lmk if you need an explanation on how to :)

[haras-unicorn]

i have this in my config so i just copied it over from there and it works for me but someone elses confirmation would be nice :)

      environment.systemPackages = with pkgs; [
        (gpt4all-cuda.overrideAttrs (final: prev: {
          nativeBuildInputs = (prev.nativeBuildInputs or [ ]) ++ [
            pkgs.autoAddDriverRunpath
          ];
        }))
      ];

[VeilSilence]

@VeilSilence Since you reported the original issue. Do you have a sec to test this? Lmk if you need an explanation on how to :)

{
  pkgs ?
    import (fetchTarball {
      url = "https://github.com/haras-unicorn/nixpkgs/archive/06c5b53ca56092bc65c2f6f30e31c8eeb87cbb7e.tar.gz";
    }) {
      config = {
        allowUnfree = true;
      };
    },
}:
pkgs.mkShell {
  buildInputs = [pkgs.gpt4all-cuda];
}

I made this shell, and build/launched gpt4all inside shell. gpt4all see cuda device, so if my test is correct, this merge request indeed fix this issue.

[Titaniumtown]

Ok given the feedback I'll approve it. I have no qualms about the changes :)

[Titaniumtown]

LGTM. others said it fixed their issues.