4

Problem: Python 3.7 is outdated, yet prebuilt containers for Vertex AI Workbench are all are built w/ 3.7.

Is there a way to modify (e.g. rebuild the base containers that work w/ Vertex AI workbench), e.g. perhaps by setting an ARG command to rebuild a working base container with Python >= 3.8?

Alternatively, does any code exist to demonstrate building an image compatible with Workbench?

Brian Bien
  • 723
  • 8
  • 21
  • Here's the best alternative I've found, but I haven't tested this yet: https://cloud.google.com/vertex-ai/docs/workbench/managed/custom-container#requirements – Brian Bien Mar 07 '23 at 14:25

2 Answers2

0

As @Brian Bien mentioned in the comments, and per my R&D the best alternative as of now is to use a Custom Container.

Posting the answer as community wiki for the benefit of the community that might encounter this use case in the future.

Feel free to edit this answer for additional information.

Scott B
  • 2,516
  • 1
  • 8
  • 15
0

GCP's latest workbench container now arrives with a later version of python.

Brian Bien
  • 723
  • 8
  • 21