Module import doesn't work when using pip install llama-cpp-python --target="dir"

When using the latest update with how I've setup my embedded install, I ran into this issue:
```
No module named 'llama_cpp.llama_chat_format'
Traceback (most recent call last):
File "Z:\Anyboty_Client\llamacppClient.py", line 252, in
llama_cpp = importlib.import_module(llama_module_dir, package="llama_cpp") # Import the llama module from the llama-cpp-python module
File "", line 1050, in _gcd_import
File "", line 1027, in _find_and_load
File "", line 1006, in _find_and_load_unlocked
File "", line 688, in _load_unlocked
File "", line 883, in exec_module
File "", line 241, in _call_with_frames_removed
File "Z:\Anyboty_Client\PythonFiles\llama_cpp_versions\12_1\ggufv2\windows10\llama_cpp\__init__.py", line 2, in
from .llama import *
File "Z:\Anyboty_Client\PythonFiles\llama_cpp_versions\12_1\ggufv2\windows10\llama_cpp\llama.py", line 27, in
import llama_cpp.llama_chat_format as llama_chat_format
ModuleNotFoundError: No module named 'llama_cpp.llama_chat_format'
```

I resolved the issue in my specific setup by changing this:
https://github.com/abetlen/llama-cpp-python/blob/01846a76b980224609ac8b69c8eb3f5785816752/llama_cpp/llama.py#L26-L27

To this:
```
from . import llama_cpp
from . import llama_chat_format
```

And while I think this is probably a fine solution, not 100% sure. So thought I'd just bring it to y'alls attention and let you fix it properly.