https://github.com/blake-hamm/kube-ai-stack
This is a helm chart I created and maintain to enable an ‘all-in-one’ ai/ml platform on kubernetes. Inspired by the kube-prometheus-stack , it includes subcharts for major LLMOps and MLOps frameworks like MLflow and Arise Phoenix . I began building this to fullfill a need in my homelab to run foundation models on limited hardware with llama.cpp and provide an ai gateway with LiteLLM . I have designed a unique way to scale models to zero using Prometheus and KubeElasti , allowing access to more models than hardware allows.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.