bugfix/issue-277_hello_world

- Added recognition of Kraken Point (KRK) APU.
- Changed torch.onnx.export inputs to dummy_input because torch
  2.9.1 does not use a dictionary.
- Finished the iteration loop over NPU and CPU implementations. Not
  a good test.  Really need a large model to show improvement of
  using the NPU.
- Added onnxscript to requirements.txt
- Mentioned the state file when cloning the conda environment.

Author: kfreidank

tutorial/hello_world/README.md

@@ -22,7 +22,9 @@ This folder contains the following files:
 
 Before running this example, ensure that you have followed the Ryzen AI Installation instructions found [here](https://ryzenai.docs.amd.com/en/latest/inst.html) and have activated the conda environment created during installation.
 
-Create a clone of the Ryzen AI installation conda environment to add required python packages
+Create a clone of the Ryzen AI installation conda environment and activate it.
+If using conda, don't forget to copy the conda-meta/state file from the Ryzen AI installation environment to your new environment's conda-meta directory.  The "conda create --clone" option does not make the copy for you. The state file setsthe RYZEN_AI_CONDA_ENV_NAME environment variable for you when you activate the conda environment.
+Add the required python packages.
 
 ```python
 set RYZEN_AI_CONDA_ENV_NAME=ryzen-ai-<version>