[DEPRECATED] Continuous Training with TensorFlow, PyTorch, XGBoost, and Scikit Learn Models with Kubeflow and AI Platform Pipelines Reviews

jupyter spawn error, please update the lab

Yao Jing Q. · Reviewed almost 2 years ago

Andrew T. · Reviewed almost 2 years ago

Gaurav R. · Reviewed almost 2 years ago

Terminal did not work. Notebook instance did not start

Balaji R. · Reviewed almost 2 years ago

Terminal did not work.

Balaji R. · Reviewed almost 2 years ago

outdated!

Meng Oon L. · Reviewed almost 2 years ago

Max S. · Reviewed almost 2 years ago

Vikash K. · Reviewed almost 2 years ago

JORGE P. · Reviewed almost 2 years ago

It is not clear why the last step didn't work. It's quite frustrating to be honest

Hector C. · Reviewed almost 2 years ago

Riccardo F. · Reviewed almost 2 years ago

Joel A. · Reviewed almost 2 years ago

saro k. · Reviewed almost 2 years ago

.

Ignacio B. · Reviewed almost 2 years ago

No able to create a Workbench Instance because there are no network available

Jorge P. · Reviewed almost 2 years ago

Lab is not working at all not able to create Vertex AI Jupyter notebook

Shailesh D. · Reviewed almost 2 years ago

Alexander G. · Reviewed almost 2 years ago

It doesnt work

Héctor O. · Reviewed almost 2 years ago

Nishana H. · Reviewed almost 2 years ago

.

Ignacio B. · Reviewed almost 2 years ago

Notebooks never get created. Wait forever until lab ends. This entire module is a heap of burning garbage.

Brian A. · Reviewed almost 2 years ago

Alexander G. · Reviewed almost 2 years ago

Better than last version but still lots of errors on this lab, first is that debian 10 images are deprecated and you should use 11 (user managed instances of Debian 10 are no longer an option), if you set up everything except Debian version you are good to go and continue with the tasks. Further on environment setup, the call for "Iterable" should came from collections.abc and never from collections, you need to edit it manually (use nano on /opt) in the last 3 steps of pipeline deploy or else an error will prevent you to complete it, lastly, the pipelines or runs are not on Vertex nor on IA Platform interface, so you should not be able to copy the ENDPOINT from there as the first lab script tell you to do (but you could use it since you manually set the ENDPOINT on prior tasks, this works but you got no green check doing so), so, you can complete all the 2 last taks on your endpoint but never got green checks for those since I think they are not on scope of where they should, also, the runs will fail on 2 first attempts cause of racing conditions (one on BQ, other on scikit) and you need to rerun those pipes (20-30 minutes task if you are fast), after that, a workaround for the last 2 checks is to manually create an instance on IA platform and use this endpoint on the last tasks.

Andre C. · Reviewed almost 2 years ago

Soway C. · Reviewed almost 2 years ago

Has error for the last two checking progress

Chua X. · Reviewed almost 2 years ago