Debug log shows: cli.azure.cli.core.extension: Unable to get extension experimental/preview status

> ### `az feedback` auto-generates most of the information requested below, as of CLI version 2.0.62

**Describe the bug**
We are developing an extension for azure machine learning. It works fine, but while running a command with --debug I see a lot of messages like below:
cli.azure.cli.core: Loading extensions:
cli.azure.cli.core: Name Load Time Groups Commands Directory
cli.azure.cli.core.extension: Unable to get extension preview status: Traceback (most recent call last):
File "C:\Users\banide\Miniconda3\envs\py222\lib\site-packages\azure\cli\core\extension\__init__.py", line 99, in preview
self._preview = bool(self.metadata.get(EXT_METADATA_ISPREVIEW))
AttributeError: 'NoneType' object has no attribute 'get'

cli.azure.cli.core.extension: Unable to get extension experimental status: Traceback (most recent call last):
File "C:\Users\banide\Miniconda3\envs\py222\lib\site-packages\azure\cli\core\extension\__init__.py", line 112, in experimental
self._experimental = bool(self.metadata.get(EXT_METADATA_ISEXPERIMENTAL))
AttributeError: 'NoneType' object has no attribute 'get'

cli.azure.cli.core.extension: Unable to get extension preview status: Traceback (most recent call last):
File "C:\Users\banide\Miniconda3\envs\py222\lib\site-packages\azure\cli\core\extension\__init__.py", line 99, in preview
self._preview = bool(self.metadata.get(EXT_METADATA_ISPREVIEW))
AttributeError: 'NoneType' object has no attribute 'get'

cli.azure.cli.core.extension: Unable to get extension experimental status: Traceback (most recent call last):
File "C:\Users\banide\Miniconda3\envs\py222\lib\site-packages\azure\cli\core\extension\__init__.py", line 112, in experimental
self._experimental = bool(self.metadata.get(EXT_METADATA_ISEXPERIMENTAL))
AttributeError: 'NoneType' object has no attribute 'get'

before eventually it gets to below after 108 times,

cli.azure.cli.core: machinelearningservices 0.377 10 54 C:\Users\banide\gitrepos\sdk-cli-v2\src\cli\src\machinelearningservices
cli.azure.cli.core: Total (1) 0.377 10 54
cli.azure.cli.core: Loaded 10 groups, 54 commands.
cli.azure.cli.core: Found a match in the command table.
cli.azure.cli.core: Raw command : ml compute show
cli.azure.cli.core: Command table: ml compute show
cli.knack.cli: Event: CommandInvoker.OnPreCommandTableTruncate []
az_command_data_logger: command args: ml compute show -n {} --debug

If this is the normal flow, I am worried , we could be wasting a some time in this stack flow. How to get rid of this flow?

for all our commands we have the command group like this:

with self.command_group("ml model", client_factory=cf_ml_cl, **is_experimental=True**) as g:

Is this not sufficient to make it experimental....?

**To Reproduce**

**Expected behavior**

**Environment summary**

**Additional context**