In this post, we explored the integration of NFS to enhance scalability in deploying LLM models within a home lab. Setting up NFS involved connecting to a TerraMaster NAS, and the K3s cluster was configured to dynamically provision storage. With NFS in place, the deployment process became more efficient, eliminating the need to rebuild images for each new model introduction. The post detailed the setup steps, from configuring NFS and K3s to deploying LLM models dynamically. This approach simplifies the scaling of Language Models, providing a centralized and scalable storage solution through NFS in a Kubernetes environment.