kthena
https://github.com/volcano-sh/kthena
Go
Lightweight, Modular, Kubernetes-native AI serving platform for scalable model serving.
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 kthena
Help out
- Issues
- benchmark: empirically calibrate all 8 router-benchmark scenarios and detect request-level saturation
- Support ModelServing revision history and rollback
- router: complete a full kvcache GC pass per tick
- router: charge rate limit quota only after a backend is selected
- feat: support maxUnavailable for ModelServing eviction protection
- router: wait for request draining before exiting on SIGTERM
- Performance: Improve the e2e perf of router
- 📢 Notice to Contributors: Guidelines for AI-Assisted Contributions
- fix: serve webhook TLS from the in-memory certificate
- perf: remove Redis writes from deny path and read-only Tokens() in GlobalRateLimiter
- Docs
- Go not yet supported