How do we set spark_version in cluster policies to select the latest GPU ML LTS version as defaultValue?

[Ramana]

Currently, I use the below two different JSON snippets to choose either Standard or ML runtime. Similar to the below, what is the defaultValue for spark_version to select the latest GPU ML LTS runtime version?

 "spark_version": {

  "type": "regex",

  "pattern": "[1-9][0-9]\\.[0-9].*",

  "defaultValue": "auto:latest-ml"

 }

 "spark_version": {

  "type": "regex",

  "pattern": "[1-9][0-9]\\.[0-9].*",

  "defaultValue": "auto:latest-lts"

[LandanG · Accepted solution]

Hi @Ramana Kancharana​ ,

As of right now these options are only available for non-GPU DBRs

[Ramana]

Thank you. Let me know once it is available.

[LandanG]

@Ramana Kancharana​ for sure, will do.