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
- chore(deps): bump sentencepiece from 0.2.2 to 0.2.3 in /cmd/runtimes/deepspeed
- chore(deps): bump kubernetes from 37.0.1 to 37.0.2 in /cmd/initializers/dataset
- fix(api): reject Torch/JAX/XGBoost runtimes whose trainer replicatedJob isn't named node. Fixes #4422
- feat(initializer): implement vLLM-style model loading fallback
- docs: update installation and Data Cache guides to v2.4.0
- fix(docs): use v2.4.0 in the installation and Data Cache guides
- Torch/JAX/XGBoost runtime whose trainer job isn't named `node` is accepted but hangs
- chore(ci): bound hung E2E notebooks and dump local container state
- chore(test): cover status server validation
- TrainJob is failed immediately when activeDeadlineSeconds exceeds 9223372036
- Docs
- Go not yet supported