This repository contains Helm charts for deploying various applications on Kubernetes.
- aibrix - A Helm chart for deploying AIBrix, a high-performance AI inference engine.
- nvidia-gpu-exporter - A Helm chart for deploying the NVIDIA GPU Exporter, which exposes GPU metrics to Prometheus.
- kuberay-dashboard - A Helm chart for deploying the Ray Dashboard, a web UI for monitoring and managing Ray clusters.
- vllm - A Helm chart for deploying vLLM, a high-throughput and memory-efficient inference engine for LLMs.
helm repo add danchev https://danchev.github.io/helm-charts
helm repo updateTo install the AIBrix chart:
helm install aibrix danchev/aibrix -n aibrix-system --create-namespaceTo install the NVIDIA GPU Exporter chart:
helm install nvidia-gpu-exporter danchev/nvidia-gpu-exporterTo install the Ray Dashboard chart:
helm install kuberay-dashboard danchev/kuberay-dashboardReplace vllm, nvidia-gpu-exporter, kuberay-dashboard with your desired release names.
For more details, see the README in each chart directory.