Using custom containers with AI Platform Training Reviews

Andrew R. · Reviewed over 2 years ago

Alikaou T. · Reviewed over 2 years ago

Marc D. · Reviewed over 2 years ago

Marc D. · Reviewed over 2 years ago

Dimas M. · Reviewed over 2 years ago

Mamoutou F. · Reviewed over 2 years ago

Anil G. · Reviewed over 2 years ago

errors where in code

Adarsha S. · Reviewed over 2 years ago

Stephanie C. · Reviewed over 2 years ago

Shantanu D. · Reviewed over 2 years ago

The Notebook seems pretty messy. For instance, the below import is never used : from kfp.components import func_to_container_op Also, quite unclear why we create a Vertex AI Pipeline and a cluster. It isn’t explicitly called anywhere. The submitted jobs have the --scale-tier="BASIC" set, which states 1 only most basic VM. There doesn’t seem to exist any way to specify an existing cluster for a `gcloud ai-platform jobs submit training $JOB_NAME` job to use. Monitoring on the created cluster even shows zero activity. full of out-of-date instructions.

Aurélien-Morgan C. · Reviewed over 2 years ago

Matías M. · Reviewed over 2 years ago

vas k. · Reviewed over 2 years ago

김기현 S. · Reviewed over 2 years ago

Mostafa S. · Reviewed over 2 years ago

Marco T. · Reviewed over 2 years ago

Phát P. · Reviewed over 2 years ago

Luis Miguel G. · Reviewed over 2 years ago

Luca D. · Reviewed over 2 years ago

Rachel B. · Reviewed over 2 years ago

Hussain B. · Reviewed over 2 years ago

Marc B. · Reviewed over 2 years ago

It is recommended to update the code to include Vertex AI training and prediction. There is a type in the description: "serch" should be "search"

Ehsan V. · Reviewed over 2 years ago

charanya p. · Reviewed over 2 years ago