Skip to content

feat(llama3.1-8b): migrate MaxText training workload to gcluster#253

Draft
kvenkatachala333 wants to merge 1 commit into
AI-Hypercomputer:mainfrom
kvenkatachala333:gcluster-migration-maxtext-v6e-8
Draft

feat(llama3.1-8b): migrate MaxText training workload to gcluster#253
kvenkatachala333 wants to merge 1 commit into
AI-Hypercomputer:mainfrom
kvenkatachala333:gcluster-migration-maxtext-v6e-8

Conversation

@kvenkatachala333

Copy link
Copy Markdown
  • Refactors the workload execution template to support Cluster Toolkit's gcluster framework.
  • Parameterizes GKE cluster and slice properties dynamically via environment variables.
  • Adds declarative kueue-resources.yaml defining topology resource flavors for v6e queue scheduling.
  • Updates README.md with comprehensive migration setup guides, useful CLI operations commands, and detailed compile-time OOM/IAM authorization troubleshooting mitigations.

Logs: https://paste.googleplex.com/5888432009248768#l=1

- Refactors the workload execution template to support Cluster Toolkit's `gcluster` framework.
- Parameterizes GKE cluster and slice properties dynamically via environment variables.
- Adds declarative `kueue-resources.yaml` defining topology resource flavors for v6e queue scheduling.
- Updates `README.md` with comprehensive migration setup guides, useful CLI operations commands, and detailed compile-time OOM/IAM authorization troubleshooting mitigations.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant