stuck in install LLAMA-CPP - windows

[kalle07]

iam only a litle in knowing of python and so on (but i get run some github repositorys like privateGPT)

OK, the server runs and the standard model responds, but only CPU....

so step by step, what and where shoudl i doo install lama.cpp (powershell, cmd, anaconda ???)

CMAKE already responds
cmake_args (dont work)
ok in know Environment Variables, but what should i write there ?

and where should i write this line

CMAKE_ARGS="-DLLAMA_CUBLAS=on" FORCE_CMAKE=1 pip install llama-cpp-python==0.1.83 --no-cache-dir