mldesigner no support beyond 3.8 and no documentation

## Which example? Describe the issue
Tried to create a new conda env with the following yaml, it works fine until python 3.8 but not with 3.9 and 3.10.
Also is there any documentation how to use it. Or, is it currently supported or not.
```
name: imagekeras_train_conda_env
channels:
- defaults
dependencies:
- python=3.8
- pip=20.0
- pip:
- mldesigner
- azureml-mlflow
```
example:
description:

## Additional context

-

Start with the supplied conda environment YAML and reproduce installation with Python 3.8, 3.9, and 3.10, including mldesigner and azureml-mlflow. Determine the supported Python versions and the intended mldesigner usage; done means the compatibility result and usage guidance are documented.

Written by the indexing model from the issue text.