trainer
https://github.com/kubeflow/trainer
Go
Distributed AI Model Training and LLM Fine-Tuning on Kubernetes
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Go not yet supported1 Subscribers
Add a CodeTriage badge to trainer
Help out
- Issues
- feat: KEP-3562: Add Phase 2 Kubeflow SDK integration design
- fix(framework): derive process counts from the resolved node resources
- fix(api): reject OptimizationJob trainer env that trial TrainJobs cannot accept
- Support user‑specified initializers in TrainJob when runtime has none
- data_cache: explore initializing cache from trainer using plugin approach
- Add controller-level Prometheus metrics and ServiceMonitor
- KEP-3562: [OptimizationJob] Integrate Kueue using `suspend` and `managedBy` fields for queue management
- chore(test): run a real OptimizationJob search without mocks
- feat: add validation for ReplicatedJobPatch name
- feat(api): support multi-objective optimization in OptimizationJob
- Docs
- Go not yet supported