Step 2: Providing Amazon Transcribe with data permissions - Amazon Transcribe

Step 2: Providing Amazon Transcribe with data permissions

To create a custom language model, you need to give Amazon Transcribe access to your text data. To do this, give Amazon Transcribe the GetObject and ListBucket permissions to your Amazon Simple Storage Service (Amazon S3) bucket.

To provide data permissions

  1. Sign in to AWS Management Console and open the IAM console at IAM console.

  2. In the navigation pane, for Access management, choose Roles and choose one of the following options.

    • Choose the name of the IAM role to use to create your custom language model and run transcription jobs.

    • To create a new IAM role, choose Create role.

      1. Under Common use cases, choose EC2. You can select any use case, but EC2 is one of the most straightforward ones.

      2. Choose Next: Tags.

      3. Choose Next: Review.

      4. Specify a name for the role under Role name. Remove the text under Role description.

      5. Choose Create role.

      6. Choose the role you've created.

  3. Choose Trust relationships.

  4. Choose Edit trust relationship.

  5. If you're using an existing role, modify your trust policy with the following code. If you've created a new role using this procedure, replace the trust policy text with the following code.

    { "Version": "2012-10-17", "Statement": [ { "Effect": "Allow", "Principal": { "Service": "" }, "Action": "sts:AssumeRole" } ] }
  6. In the navigation pane, choose Policies.

    • From the policy list, choose AmazonTranscribeFullAccess.

      1. From Policy actions, choose Attach policy

      2. Choose the IAM role that you want to attach the policy to.

    • Choose Create policy.

      1. Choose JSON.

      2. Enter the following code, which adds GetObject and ListBucket permissions:

        { "Version": "2012-10-17", "Statement": [ { "Action": [ "s3:GetObject" ], "Resource": [ "arn:aws:s3:::your-input-bucket/*" ], "Effect": "Allow" }, { "Action": [ "s3:ListBucket" ], "Resource": [ "arn:aws:s3:::your-input-bucket" ], "Effect": "Allow" } ] }
      3. Choose Review policy.

      4. Specify a name for the policy.

      5. Choose Create policy.

      6. Attach the policy to the IAM role.

Amazon Transcribe now has the necessary permissions to access the data for your custom language model.

Next step

Step 3: Creating a custom language model