From ca6ec9b37069f2d3f76b67116f3bf3e586b7e60a Mon Sep 17 00:00:00 2001
From: Dan Garfield <dan@codefresh.io>
Date: Fri, 18 Aug 2023 10:48:24 -0600
Subject: [PATCH] Add Kubernetes support (#18)

---
 deploy/kubernetes/kustomization.yaml          | 20 +++++++++++++
 .../kubernetes/llama-gpt-api-deployment.yaml  | 29 +++++++++++++++++++
 deploy/kubernetes/llama-gpt-api-service.yaml  | 15 ++++++++++
 .../kubernetes/llama-gpt-ui-deployment.yaml   | 26 +++++++++++++++++
 deploy/kubernetes/llama-gpt-ui-service.yaml   | 16 ++++++++++
 5 files changed, 106 insertions(+)
 create mode 100644 deploy/kubernetes/kustomization.yaml
 create mode 100644 deploy/kubernetes/llama-gpt-api-deployment.yaml
 create mode 100644 deploy/kubernetes/llama-gpt-api-service.yaml
 create mode 100644 deploy/kubernetes/llama-gpt-ui-deployment.yaml
 create mode 100644 deploy/kubernetes/llama-gpt-ui-service.yaml

diff --git a/deploy/kubernetes/kustomization.yaml b/deploy/kubernetes/kustomization.yaml
new file mode 100644
index 0000000..72daa6f
--- /dev/null
+++ b/deploy/kubernetes/kustomization.yaml
@@ -0,0 +1,20 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+
+resources:
+- llama-gpt-api-deployment.yaml
+- llama-gpt-api-service.yaml
+- llama-gpt-ui-deployment.yaml
+- llama-gpt-ui-service.yaml
+
+# patches:
+# - 
+
+configMapGenerator:
+- name: llama-gpt
+  literals:
+  - DEFAULT_MODEL="/models/llama-2-7b-chat.bin"
+  - OPENAI_API_HOST="http://llama-gpt-api:8000"
+  - OPENAI_API_KEY="sk-XXXXXXXXXXXXXXXXXXXX"
+  - WAIT_HOSTS="llama-gpt-api:8000"
+  - WAIT_TIMEOUT="600"
\ No newline at end of file
diff --git a/deploy/kubernetes/llama-gpt-api-deployment.yaml b/deploy/kubernetes/llama-gpt-api-deployment.yaml
new file mode 100644
index 0000000..47a0736
--- /dev/null
+++ b/deploy/kubernetes/llama-gpt-api-deployment.yaml
@@ -0,0 +1,29 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  labels:
+    service: llama-gpt-api
+  name: llama-gpt-api
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      service: llama-gpt-api
+  template:
+    metadata:
+      labels:
+        service: llama-gpt-api
+    spec:
+      containers:
+        - name: llama-gpt-api
+          image: ghcr.io/getumbrel/llama-gpt-api:1.0.1
+          env:
+            - name: MODEL
+              valueFrom: 
+                configMapKeyRef:
+                  name: llama-gpt
+                  key: DEFAULT_MODEL
+          resources:
+            requests:
+              memory: 5Gi
+      restartPolicy: Always
\ No newline at end of file
diff --git a/deploy/kubernetes/llama-gpt-api-service.yaml b/deploy/kubernetes/llama-gpt-api-service.yaml
new file mode 100644
index 0000000..ac8a3bd
--- /dev/null
+++ b/deploy/kubernetes/llama-gpt-api-service.yaml
@@ -0,0 +1,15 @@
+apiVersion: v1
+kind: Service
+metadata:
+  labels:
+    service: llama-gpt-api
+  name: llama-gpt-api
+spec:
+  ports:
+    - name: api
+      port: 8000
+      targetPort: 8000
+  selector:
+    service: llama-gpt-api
+status:
+  loadBalancer: {}
diff --git a/deploy/kubernetes/llama-gpt-ui-deployment.yaml b/deploy/kubernetes/llama-gpt-ui-deployment.yaml
new file mode 100644
index 0000000..c4bd001
--- /dev/null
+++ b/deploy/kubernetes/llama-gpt-ui-deployment.yaml
@@ -0,0 +1,26 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  labels:
+    service: llama-gpt-ui
+  name: llama-gpt-ui
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      service: llama-gpt-ui
+  template:
+    metadata:
+      labels:
+        service: llama-gpt-ui
+    spec:
+      containers:
+        - name: llama-gpt-ui
+          image: ghcr.io/getumbrel/llama-gpt-ui:latest
+          envFrom:
+          - configMapRef:
+              name: llama-gpt
+          ports:
+            - containerPort: 3000
+          resources: {}
+      restartPolicy: Always
\ No newline at end of file
diff --git a/deploy/kubernetes/llama-gpt-ui-service.yaml b/deploy/kubernetes/llama-gpt-ui-service.yaml
new file mode 100644
index 0000000..199eb78
--- /dev/null
+++ b/deploy/kubernetes/llama-gpt-ui-service.yaml
@@ -0,0 +1,16 @@
+apiVersion: v1
+kind: Service
+metadata:
+  labels:
+    service: llama-gpt-ui
+  name: llama-gpt-ui
+spec:
+  ports:
+    - name: ui
+      port: 3000
+      targetPort: 3000
+  selector:
+    service: llama-gpt-ui
+  type: ClusterIP
+status:
+  loadBalancer: {}