Build does not use correct clang

I downloaded and built the model (microsoft/BitNet-b1.58-2B-4T-gguf) using my Intel Mac following the instructions on README.
I installed llvm using brew.

$ which clang
/usr/local/opt/llvm/bin/clang
$ clang --version
Homebrew clang version 20.1.3
Target: x86_64-apple-darwin24.4.0
Thread model: posix
InstalledDir: /usr/local/Cellar/llvm/20.1.3/bin
Configuration file: /usr/local/etc/clang/x86_64-apple-darwin24.cfg

I added the following commands to ~/.zshrc

export PATH="/usr/local/opt/llvm/bin:$PATH"
export LDFLAGS="-L/usr/local/opt/llvm/lib"
export CPPFLAGS="-I/usr/local/opt/llvm/include"
export CC=clang
export CXX=clang++
export LD=ld.lld
export AR=llvm-ar
export RANLIB=llvm-ranlib

However, when I run run_inference.py, it tells me that the model was built using the native clang on my machine (which is a lower version than the requirement clang>=18):

build: 3957 (5eb47b72) with Apple clang version 17.0.0 (clang-1700.0.13.3) for x86_64-apple-darwin24.4.0

Does anyone know how to fix this?

Start with the README build instructions and the compiler invocation reached through run_inference.py. Trace why the build reports Apple clang 17 despite the configured Homebrew LLVM environment, then verify that the model reports the intended clang version after rebuilding.