Using Kubeflow Pipelines with AI Platform Reviews

Claudiu-Alexandru D. · Reviewed מעל 4 שנים ago

There are still errors in the pipelines inside Jupyter Notebook

Fidelis E. · Reviewed מעל 4 שנים ago

Shazad Khalid K. · Reviewed מעל 4 שנים ago

ccz c. · Reviewed מעל 4 שנים ago

Saurav K. · Reviewed מעל 4 שנים ago

MBTKI M. · Reviewed מעל 4 שנים ago

Nilimesh H. · Reviewed מעל 4 שנים ago

Andrew B. · Reviewed מעל 4 שנים ago

Amor Y. · Reviewed מעל 4 שנים ago

Dan M. · Reviewed מעל 4 שנים ago

Muhammad A. · Reviewed מעל 4 שנים ago

Nqaba N. · Reviewed מעל 4 שנים ago

Provisioned project didn't have required API (AI Platform Training & Prediction API) enabled, which caused runs to fail. I was able to look in the error logs to diagnose this problem; but that's no excuse - the project should have had this API enabled. A number of other users are complaining that nothing is working, and I'd be willing to bet that this is the reason why. Also, the instructions could use some editing: Task 4, Step 1 tells us to create a bucket in the console; but Step 2 tells us to do the same thing in the terminal window. This is quite obviously a mistake. Finally, in Task 5, Step 3, you don't tell us to comment out the first call to `kfp.Client().create_run_from_pipeline_func()`. I know the example below has this commented out, but it should also be mentioned in the instructions.

Matthew J. · Reviewed מעל 4 שנים ago

Florian M. · Reviewed מעל 4 שנים ago

Tim T. · Reviewed מעל 4 שנים ago

Jhoan N. · Reviewed מעל 4 שנים ago

Vishwanath P. · Reviewed מעל 4 שנים ago

Thad C. · Reviewed מעל 4 שנים ago

Manuel F. · Reviewed מעל 4 שנים ago

Grégoire M. · Reviewed מעל 4 שנים ago

Shubham K. · Reviewed מעל 4 שנים ago

nothing working

Kalinkin A. · Reviewed מעל 4 שנים ago

Oscar K. · Reviewed מעל 4 שנים ago

g k. · Reviewed מעל 4 שנים ago

Alfin P. · Reviewed מעל 4 שנים ago