Tagging with AWS DeepComposer

A tag is label that you or AWS optionally defines and associates with AWS resources, including certain types of AWS DeepComposer resources and actions.

In the AWS DeepComposer service, you can tag models trained in the AWS DeepComposer console and compositions you have created using AWS DeepComposer music studio.

For more information about using tags, use Tagging best practices.

Managing Tags

A tag key can contain as many as 128 characters. A tag value can contain as many as 256 characters. The characters can be Unicode letters, numbers, white space, or one of the following symbols: _ . : / = + -. The following additional restrictions apply to tags:

• Tag keys and values are case sensitive.

• For each associated resource, each tag key must be unique and have only one value.

• Tags are available only for your AWS account, not any other accounts that share the resource. In addition, the tags are available only for resources that are located in the specified AWS Region for your AWS account.

• Do not use aws:, AWS: or any upper or lowercase combination of such as a prefix for keys, because it is reserved for AWS use. You cannot edit or delete tag keys with this prefix. Values can have this prefix. If a tag value has aws as its prefix but the key does not, then AWS DeepComposer considers it to be a user tag and it counts against the limit of 50 tags. Tags with only the key prefix of aws do not count against your tags-per-resource limit.

You can use list below to find more information about creating and managing tags for either models or compositions created in AWS DeepComposer.

• Creating and editing tags for models trained in AWS DeepComposer

• Creating and editing tags for compositions you have created in the AWS DeepComposer music studio.

Using tags in IAM policies

After you start implementing tags, you can apply tag-based, resource-level permissions to AWS Identity and Access Management (IAM) policies. This includes operations that support adding tags to resources when resources are created. By using tags in this way, you can implement granular control of which groups and users in your AWS account have permission to access resources in AWS DeepComposer.

{
   "Version": "2012-10-17",
   "Statement": [
      {
         "Sid": "VisualEditor1",
         "Effect": "Allow",
         "Action": [
            "deepcomposer:GetComposition",
            "deepcomposer:UntagResource"
         ]
         "Resource": "*",
         "Condition": {
            "StringEquals: {
               "aws:ResourceTag/tagKey": "tagValue"
            }
         }
      }
   ]
}
        

If you define tag-based, resource-level permissions, the permissions take effect immediately. This means that your resources are more secure as soon as they're created, and you can quickly start enforcing the use of tags for new resources. You can also use resource-level permissions to control which tag keys and values can be associated with new and existing resources. For more information, see Controlling access using tags in the AWS IAM User Guide.

Creating and editing tags for models trained in AWS DeepComposer

You can use the procedures below to create and manage tags for models in the AWS DeepComposer console.

Adding a tag while training a model in the AWS DeepComposer console

1. Open the AWS DeepComposer console.

2. In the navigation pane, choose Models.

3. On the Models page, choose Create a model.

4. On the Train a model page, under Generative algorithm, choose an available option.

5. Choose an available Training dataset.

6. (Optional) Under Hyperparameters, update the default values to adjust how your custom model is trained.

7. Under Model details, give your model a name and an optional description.

8. Under Tags, choose Add a new tag.

9. For Key and Value, enter appropriate values.

   For example, BachModels and Model01, respectively.

10. (Optional) To add another tag, choose Add new tag.

11. (Optional) To remove a tag, choose Remove.

12. Choose Start training to begin training your model.

After tagging and submitting your new model for training, you can manage its tags during training or after training has completed. Use the procedure below to manage existing tags.

Managing tags for AWS DeepComposer models

1. Open the AWS DeepComposer console.

2. In the navigation pane, choose Models.

3. On the Models page, choose your model.

4. Choose View details.

5. Under Tags, choose Manage tags.

6. (Optional) To remove an existing tag, choose Remove.

7. (Optional) To add a new tag, choose Add new tag.

8. For Key and Value, enter appropriate values.

   For example, PopModels and Model01, respectively.

9. When finished managing your tags for that specific model, choose Save changes.

Creating and editing tags for compositions created in the AWS DeepComposer music studio

You can create and edit tags for compositions you have created in the AWS DeepComposer music studio. Unlike models, you can only create and edit tags after you have generated the composition in the AWS DeepComposer music studio.

Managing tags for AWS DeepComposer compositions

1. Open the AWS DeepComposer console.

2. In the navigation pane, choose Compositions.

3. On the Compositions page, choose you composition.

4. In the Actions drop down menu, choose Manage tags.

5. To add a new tag, choose Add new tag.

6. For Key and Value, enter appropriate values.

   For example, JazzModel and Model01, respectively.

7. (Optional) To remove an existing tag, choose Remove.

8. When finished managing your tags for that specific composition, choose Save.

Additional information

For more information about tagging, see the following resources.

• AWS Tagging Principles in the AWS General Reference

• AWS Tagging Strategies (downloadable PDF)

• AWS Access Control in the AWS IAM User Guide

• AWS Tagging Policies in the AWS Organizations User Guide