Make run model a oneshot service

Signed-off-by: baalajimaestro <me@baalajimaestro.me>

Dockerfile

@@ -25,18 +25,19 @@ gpgkey=https://yum.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.
 
 RUN mv /tmp/oneAPI.repo /etc/yum.repos.d/oneAPI.repo && \
     dnf update && \
-    dnf install intel-oneapi-mkl -y && \
+    dnf install intel-oneapi-mkl epel-release -y && \
+    dnf update && \
+    dnf install supervisor -y && \
     mkdir /ollama
 
 COPY --from=build /app/ollama /ollama
-
-ADD entrypoint.sh /
+COPY --from=build /app/serve.sh /ollama/serve.sh
+COPY --from=build /app/run_model.sh /ollama/run_model.sh
+COPY --from=build /app/supervisord.conf /etc/supervisor/conf.d/supervisord.conf
 
 ENV MODEL_NAME="llama"
 ENV OLLAMA_HOST="0.0.0.0:8080"
 
 EXPOSE 8080
 
-ENTRYPOINT ["/entrypoint.sh"]
-
-CMD ["/ollama/ollama", "run", "model"]
+CMD ["supervisord", "-c", "/etc/supervisor/conf.d/supervisord.conf"]

run_model.sh

@@ -2,7 +2,7 @@
 
 . /opt/intel/oneapi/setvars.sh
 
-/ollama/ollama serve &
+sleep 10
 
 if [ -n "$INIT_MODELS" ]; then
 	/ollama/ollama rm model || true
@@ -10,4 +10,4 @@ if [ -n "$INIT_MODELS" ]; then
 	/ollama/ollama create -f /tmp/Modelfile model
 fi
 
-exec $@
+/ollama/ollama run model

serve.sh

@@ -0,0 +1,5 @@
+#! /usr/bin/env bash
+
+. /opt/intel/oneapi/setvars.sh
+
+/ollama/ollama serve

supervisord.conf

@@ -0,0 +1,14 @@
+[supervisord]
+nodaemon=true
+
+[program:ollama]
+command=/ollama/serve.sh
+autostart=true
+autorestart=true
+
+[program:run_model]
+command=/ollama/run_model.sh
+autostart=true
+autorestart=false
+startsecs=0
+exitcodes=0