aws-samples / amazon-rekognition-code-samples Goto Github PK

The Content Moderation Immersion Day workshop uses the DataScience 1.0 image (see https://github.com/aws-samples/amazon-rekognition-code-samples/blob/main/content-moderation/dist/content-moderation-im-setup.yml#L22) which was deprecated in October 2023 and is no longer shown in SageMaker Studio.

If the user tries to manually create a new app from the Data Science 3.0, it runs into an issue

due to missing sagemaker:AddTags permissions

At the end of "identity-verification/module_1" there is the IDV Demo App solution.

Creating a new user allows the use upper case letters in the username, but the login part doesn't allow the use of upper case letters for the username and thus can't login.

For "identity-verification/module_1" the first command in the SageMaker Jumpstart needs a new library. The error does provide guidance on the new library to use.

now: !pip install boto3 matplotlib opencv-python openpyxl pandas sklearn scikit-image
required: !pip install boto3 matplotlib opencv-python openpyxl pandas scikit-learn scikit-image

The error: The 'sklearn' PyPI package is deprecated, use 'scikit-learn'

For "identity-verification/module_1": At the very beginning of the lab in SageMaker Jumpstart, there is an issue with the name of the bucket.
While doing the lab I used a very common name for the S3 bucket. I got a confusing error, just mentioning that if I create that bucket in a region other than us-east-1 I should provide the region. The region was the region I actually needed and I realised after some time that I used a common name for the bucket and that was the issue. So the real issue here was not having a unique S3 bucket name.

The problem here can be solved with a more useful error message if that is possible.