🗃️ Model Deployment
3 items
📄️ Multi-Node Inference
This page describes the multi-node inference capabilities in Kthena, based on real-world examples and configurations.
📄️ Autoscaler
Overview
🗃️ Workload
3 items
🗃️ Router
12 items
🗃️ Observability
1 item
📄️ Runtime
Kthena Runtime is a lightweight sidecar service designed to standardize Prometheus metrics from inference engines, provide LoRA adapter download/load/unload capabilities, support model downloading, and publish KV cache events for the kvcache-aware router plugin — either to Redis or directly to router instances in memory push mode.
🗃️ Prefill Decode Disaggregation
4 items