llama.cpp.install


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

post_install() {
  echo ":: llama.cpp now runs as a dedicated unprivileged user."
  echo ":: NOTE: Model files must now be accessible by the llama-cpp user."
  echo ":: Previously stored models in your home folder need to be moved to:"
  echo "   /var/lib/llama-cpp/"
  echo ":: To move your existing models:"
  echo "   sudo mv ~/path/to/your/models/* /var/lib/llama-cpp"
  echo "   sudo chown -R llama-cpp:llama-cpp /var/lib/llama-cpp"
  echo ":: To allow your user to access the API socket:"
  echo "   sudo usermod -aG llama-cpp \$USER"
  echo "   (Log out and back in for group changes to apply)"
}

post_remove() {
  echo ":: The llama-cpp system user/group remains on the system."
  echo ":: To fully remove: sudo gpasswd -d \$USER llama-cpp && sudo userdel llama-cpp"
  echo ":: Remember to cleanup /var/lib/llama-cpp (contains models and data),"
  echo "   as it's not automatically removed."
  echo ":: NOTE: Running sudo userdel -r llama-cpp (with '-r' added)"
  echo "   would nuke /var/lib/llama-cpp."
}