0

I am newbie in using sagemaker pipeline .While following this AWS MLOPs workshop https://catalog.us-east-1.prod.workshops.aws/workshops/741835d3-a2bf-4cb6-81f0-d0bb4a62edca/en-US/03-lab-03 I am getting error on notebook 3 at step 67 pipeline.upsert(role_arn=sm_role) Below are the error msgs.

AttributeError: 'ParameterString' object has no attribute 'startswith'

TypeError: Pipeline variables do not support str operation. Please use .to_string() to convert it to string type in execution timeor use .expr to translate it to Json for display purpose in Python SDK.

Any help will be appreciated, thanks in advance.

Ananda Roy
  • 67
  • 1
  • 12

1 Answers1

0

Welcome to AWS!

I had this same issue. It was fixed in the new version of the Python SDK a few weeks ago.

I had to create a new user in Studio (if you're in notebook instances, create a new notebook).

Then, replace this line

! pip install "sagemaker>=2.99.0"

with this one

!{sys.executable} -m pip install 'sagemaker' --upgrade

Hope that helps!

This solution was taken from joseproura on this issue, referencing this bug fix.

Augustin Marcin
  • 21
  • 5