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
- fix(runtimes): ignore the snapshot of a deleted TrainJob on create
- Fix the statusserver auth to use offline OIDC token validation using the API server's internal discovery endpoint
- Extend MPI Plugin with MPICH and Intel Backends
- feat(mpi): add --disable-mpi-rbac-management flag
- KEP-3562: [OptimizationJob] Add Python SDK integrations for OptimizationJob in Kubeflow Trainer
- KEP-3562: [OptimizationJob] Add End-to-End (E2E) integration tests for OptimizationJob workflow
- Add GPU utilization Grafana dashboard with dcgm-exporter metrics
- feat(plugins): validate environment variable collisions across runtime plugins
- feat(api): allow spec.trainer.numNodes to be mutable when TrainJob is suspended
- fix(framework): give the Volcano PodGroup a single deliberate priority class
- Docs
- Go not yet supported