OSError on ARM (Linux/Debian) when running 'llm models' after installing llm-gpt4all

Environment:

• Operating System: Debian GNU/Linux 6.6.15-arm64
• Architecture: aarch64
• Shell: bash
• Python Version: 3.11
• pipx Version: 1.3.4
• llm Version: 0.13.1

Reproduction Steps:

1 Install llm using pipx.
2 Run the command llm install llm-gpt4all.
3 Attempt to list models using llm models.

Expected Result: The llm models command should list available models without any errors.

Actual Result: An OSError is thrown indicating a missing shared object file:

OSError: /home/user/.local/share/pipx/venvs/llm/lib/python3.11/site-packages/gpt4all/llmodel_DO_NOT_MODIFY/build/libllama.so: cannot open shared object file: No such file or directory

Additional Information: The issue seems to be specific to the ARM architecture, as the shared object file libllama.so is either not present or not correctly built for ARM.

Steps to Resolve (without gpt4all):
pipx uninstall llm
pipx install llm

Reproduce the issue on Debian arm64 with Python 3.11 by installing llm-gpt4all through pipx and running llm models. Start by inspecting how the installed gpt4all package builds or locates libllama.so; done means the command lists available models without raising the missing shared-library OSError.