NotImplementedError in update with AzureMachineLearningFileSystem

I have a problem with upload on azure
https://learn.microsoft.com/pl-pl/azure/machine-learning/how-to-access-data-interactive?view=azureml-api-2&tabs=adls
I use sample

from azureml.fsspec import AzureMachineLearningFileSystem
# instantiate file system using following URI
fs = AzureMachineLearningFileSystem('azureml://subscriptions/<subid>/resourcegroups/<rgname>/workspaces/<workspace_name>/datastores/<datastorename>/paths/')

# you can specify recursive as False to upload a file
fs.upload(lpath='data/upload_files/crime-spring.csv', rpath='data/fsspec', recursive=False, **{'overwrite': 'MERGE_WITH_OVERWRITE'})

# you need to specify recursive as True to upload a folder
fs.upload(lpath='data/upload_folder/', rpath='data/fsspec_folder', recursive=True, **{'overwrite': 'MERGE_WITH_OVERWRITE'})

This is offical solution.

When i try

from azureml.fsspec import AzureMachineLearningFileSystem
# instantiate file system using following URI
fs = AzureMachineLearningFileSystem('azureml://subscriptions/'+auth_subscription_id+'/resourcegroups/'+auth_resource_group+'/workspaces/'+auth_workspace_name+'/datastores/azure_churn_lab_datastores/paths/')

# you can specify recursive as False to upload a file
fs.upload(lpath='/' + path + seq_name + '.csv', 
          rpath='/' + path_target)

I have error with NotImplementedError:
I havent information about this problem... I tried upgrade package azure sdk.

I want ask you - this is problem with implementation or azure?
I know about different way to upload with folder.

Second question
If you prepare data in python sdk2 and default path is

You upload to blob containers.

,

or leave in this path (File Share)? And why? Because I saw - few people save and upload

I use Visual Studio Code - connect to

and when I save a file

df.to_csv('/test.csv')

default path is File Share code

> NotImplementedError: I haven't information about this problem...

The NotImplementedError occurs here while using the AzureMachineLearningFileSystem class to upload a file is most likely the result of a version conflict between the azureml-sdk and fsspec packages. To check whether it fixes the problem, try updating both packages to the latest or relevant versions.

I tried with the same code to upload a file as you have selected the default directory(File share directory).

Code:

from azureml.fsspec import AzureMachineLearningFileSystem

fs = AzureMachineLearningFileSystem('azureml://subscriptions/<SUB-ID>/resourcegroups/<RESOURCE-GROUP>/workspaces/<WORKSPACE-NAME>/datastores/workspaceworkingdirectory/paths/')

# you can specify recursive as False to upload a file
fs.upload(lpath='C:\\Users\\Downloads\\owid-covid-data.csv', rpath='data/fsspec', recursive=False, **{'overwrite': 'MERGE_WITH_OVERWRITE'})

The above code executed and uploaded the file in the default data store.

Output:

For the second question,Blob storage can be a better option if you work with a lot of unstructured data.File sharing could be the better choice if you need higher levels of access control and are working with smaller volumes of structured data.

Reference:
Access data from Azure cloud storage during interactive development - Azure Machine Learning | Microsoft Learn