TFX on Google AI LAB

I only get to 60%, i’m stuck in creating a run, it always throws a error. Maybe it is something ill specified before. Anyone with a similar issue?

I get:

kfp_server_api.exceptions.ApiException: (400)
Reason: Bad Request
HTTP response headers: HTTPHeaderDict({‘Content-Type’: ‘application/json’, ‘Date’: ‘Mon, 13 Sep 2021 02:34:40 GMT’, ‘Vary’: ‘Origin’, ‘X-Content-Type-Options’: ‘nosniff’, ‘X-Frame-Options’: ‘SAMEORIGIN’, ‘X-Powered-By’: ‘Express’, ‘X-Xss-Protection’: ‘0’, ‘Transfer-Encoding’: ‘chunked’, ‘Set-Cookie’: ‘S=cloud_datalab_tunnel=SeeYrml80va1g14xinpasYetMEPbjmaVL6FiGNBaVZE; Path=/; Max-Age=3600’})
HTTP response body: {“error”:“Failed to create a new run.: Failed to fetch workflow spec.: Invalid input error: Please provide a valid pipeline spec”,“message”:“Failed to create a new run.: Failed to fetch workflow spec.: Invalid input error: Please provide a valid pipeline spec”,“code”:3,“details”:[{"@type":“type.googleapis.com/api.Error",“error_message”:"Please provide a valid pipeline spec”,“error_details”:“Failed to create a new run.: Failed to fetch workflow spec.: Invalid input error: Please provide a valid pipeline spec”}]}

Hi Antonio! Welcome to Discourse! Have you set the ENDPOINT and RUN_ID in the notebook as shown in the instructions? It seems one of those might need to be double-checked. If it didn’t work, can you let us know which part of the notebook exactly you’re getting that error? Thanks!

The endpoint fells right. The notebook errors when creating a run, i can’t create it. Everything before works fine. In the endpoint i’m including the http:// part, should i?

SOLUTION: Read the README.md on the git repo before. My mistake, sorry.

Great! Can you let us know which part exactly was the problem? That way we can inform other learners who may run into the same error message. Thank you!

1 Like

In the README.md there is a series of instructions to execute on the console before actually start working on the notebook. Executing those solved the problem.

Got it. Thank you and hope you’ve already passed the rest of the assignment!

Can we please upgrade the text in the lab instructions to avoid people having to waste too much time figuring they have to read the README?

Hi Marjori! Welcome to Discourse and thank you for the suggestion. I believe this might be a bug because I was trying to replicate it yesterday and couldn’t. I’ve also done the lab at least 2 times before and didn’t encounter it. I was always able to create the pipeline right away. I’m thinking it might be related to the automatic creation of the Kubernetes cluster. Will add a note in the Coursera classroom about this workaround and also inform the Qwiklabs team so they can investigate. Just to be sure, are you referring to the README on this page?

Thanks again!

1 Like