Fix AzureML 101 tutorial following hands-on workshop

[jfomhover]

This adds multiple issues together to make it easier to report.

• in multiple notebooks in sdk v2 folder, or in e2e tutorials, we use DefaultAzureCredentials within an unnecessary try/catch
• Renaming "dataset" to "data asset"

[lebaro-msft]

Nit: the cell for creating the compute target kept running for 30 seconds longer than it took the toast notification to show "provisioning succeeded" which was a bit confusing, not sure what the cell was doing

[lebaro-msft]

In earlier cells in the notebook, it used type=AssetTypes.URI_FILE, but later for Input/Output it used type-"uri_file". Seems like it can reuse the constant for consistency

[jfomhover]

Using webbrowser.open() doesn't work when running in the notebooks experience, printing the URL in the cell results could help.

[jfomhover]

Looks like we can now use model="my-model@latest" to use latest version of a model. We could use that instead of the workaround to figure out the latest version.

[singankit]

Enable logging signature and example in autolog : this mentions what params can be passed to autolog> mlflow.sklearn — MLflow 1.26.1 documentation. Two of them related to signature and examples are log_input_examples, log_model_signatures. This enables logging examples and signatures of model. Latter is enable by default