diff --git a/other/vllm/Docker/Dockerfile.base b/other/vllm/Docker/Dockerfile.base
index b9ae3b8..13da536 100644
--- a/other/vllm/Docker/Dockerfile.base
+++ b/other/vllm/Docker/Dockerfile.base
@@ -30,7 +30,10 @@ RUN /venv/bin/pip3 install torch==2.0.1 --index-url https://download.pytorch.org
 
 # WORKDIR /local-llm-server
 
-RUN /venv/bin/pip install git+https://github.com/vllm-project/vllm
+# Don't build VLLM because we don't do that on the inference server. Just install from pip.
+# RUN /venv/bin/pip install git+https://github.com/vllm-project/vllm
+
+RUN pip install vllm
 
 FROM nvidia/cuda:11.8.0-base-ubuntu22.04 as base