Merge pull request #7 from defenseunicorns/create-lfai-api-zarf-package

Create LFAI-api Zarf package

.github/workflows/environment.yaml

@@ -1,5 +1,5 @@
 name: Run tests
-on: [push, pull_request]
+on: [pull_request]
 
 jobs:
   Run-Tests:

.github/workflows/publish-zarf-package.yaml

@@ -0,0 +1,40 @@
+name: Publish Artifacts 
+
+on:
+  push:
+    tags:
+      - "*"
+
+jobs:
+  zarf:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Check out code
+      uses: actions/checkout@v2
+
+    - name: Login to GHCR
+      uses: docker/login-action@v3
+      with:
+        registry: ghcr.io
+        username: ${{ github.actor }}
+        password: ${{ secrets.GITHUB_TOKEN }}
+
+    - name: Login to Registry1
+      uses: docker/login-action@v3
+      with:
+        registry: registry1.dso.mil
+        username: ${{ secrets.REGISTRY1_USERNAME }}
+        password: ${{ secrets.REGISTRY1_PASSWORD }}
+
+    - name: Build API Image
+      run: make docker-build docker-push VERSION=$GITHUB_REF_NAME
+
+    - name: Install Zarf
+      uses: defenseunicorns/setup-zarf@main
+
+    - name: Build Zarf Package
+      run: zarf package create . --confirm
+
+    - name: Publish Zarf Package
+      run: zarf package publish zarf-package-*.zst oci://ghcr.io/defenseunicorns/packages

.gitignore

@@ -2,4 +2,6 @@
 leapfrogai_api.egg-info
 __pycache__
 .vscode
-config.yaml
+config.yaml
+*.tar.zst
+.python-version

Makefile

@@ -1,4 +1,4 @@
-VERSION := $(shell git describe --abbrev=0 --tags)
+VERSION ?= $(shell git describe --abbrev=0 --tags)
 
 create-venv:
 	python -m venv .venv
@@ -16,16 +16,16 @@ build-requirements:
 	pip-compile -o requirements.txt pyproject.toml
 
 build-requirements-dev:
-	pip-compile --extra dev -o requirements-dev.txt pyproject.toml
+	pip-compile --extra dev -o requirements-dev.txt pyproject.toml --allow-unsafe
 
 test:
 	pytest **/*.py
 
 dev:
 	uvicorn main:app --port 3000 --reload
 
-make docker-build:
-	docker build -t ghcr.io/defenseunicorns/leapfrogai-api:${VERSION} .
+docker-build:
+	docker build -t ghcr.io/defenseunicorns/leapfrogai/leapfrogai-api:${VERSION} .
 
-make docker-push:
-	docker push ghcr.io/defenseunicorns/leapfrogai-api:${VERSION}
+docker-push:
+	docker push ghcr.io/defenseunicorns/leapfrogai/leapfrogai-api:${VERSION}

backends/openai/grpc_client.py

@@ -34,15 +34,16 @@ async def stream_completion(model: Model, request: leapfrogai.CompletionRequest)
 async def completion(model: Model, request: leapfrogai.CompletionRequest):
     async with grpc.aio.insecure_channel(model.backend) as channel:
         stub = leapfrogai.CompletionServiceStub(channel)
-        response: leapfrogai.CompletionResponse = stub.Complete(request)
+        response: leapfrogai.CompletionResponse = await stub.Complete(request)
 
         return CompletionResponse(
             model=model.name,
             choices=[
                 CompletionChoice(
                     index=0,
                     text=response.choices[0].text,
-                    finish_reason=response.choices[0].finish_reason,
+                    finish_reason=str(response.choices[0].finish_reason),
+                    logprobs=None,
                 )
             ],
         )
@@ -63,7 +64,7 @@ async def stream_chat_completion(
 async def chat_completion(model: Model, request: leapfrogai.ChatCompletionRequest):
     async with grpc.aio.insecure_channel(model.backend) as channel:
         stub = leapfrogai.CompletionServiceStub(channel)
-        response: leapfrogai.ChatCompletionResponse = stub.Complete(request)
+        response: leapfrogai.ChatCompletionResponse = await stub.Complete(request)
 
         return ChatCompletionResponse(
             model=model.name,

backends/openai/types.py

@@ -27,8 +27,8 @@ class CompletionRequest(BaseModel):
 class CompletionChoice(BaseModel):
     index: int
     text: str
-    logprobs: object | None
-    finish_reason: str
+    logprobs: object = None
+    finish_reason: str = ""
 
 
 class CompletionResponse(BaseModel):
@@ -37,7 +37,7 @@ class CompletionResponse(BaseModel):
     created: int = 0
     model: str = ""
     choices: list[CompletionChoice]
-    usage: Usage
+    usage: Usage = None
 
 
 ##########

chart/Chart.yaml

@@ -0,0 +1,24 @@
+apiVersion: v2
+name: leapfrogai
+description: A deployment of AI tools
+
+# A chart can be either an 'application' or a 'library' chart.
+#
+# Application charts are a collection of templates that can be packaged into versioned archives
+# to be deployed.
+#
+# Library charts provide useful utilities or functions for the chart developer. They're included as
+# a dependency of application charts to inject those utilities and functions into the rendering
+# pipeline. Library charts do not define any templates and therefore cannot be deployed.
+type: application
+
+# This is the chart version. This version number should be incremented each time you make changes
+# to the chart and its templates, including the app version.
+# Versions are expected to follow Semantic Versioning (https://semver.org/)
+version: 0.4.0
+
+# This is the version number of the application being deployed. This version number should be
+# incremented each time you make changes to the application. Versions are not expected to
+# follow Semantic Versioning. They should reflect the version the application is using.
+# It is recommended to use it with quotes.
+appVersion: "1.16.0"

chart/templates/api/deployment.yaml

@@ -0,0 +1,68 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: api-deployment
+spec:
+  replicas: {{ .Values.api.replcias }}
+  strategy:
+    type: Recreate
+  selector:
+    matchLabels:
+      app: api
+  template:
+    metadata:
+      labels:
+        app: api
+    spec:
+      serviceAccountName: read-configmaps
+      containers:
+        - name: sidecar
+          image: registry1.dso.mil/ironbank/kiwigrid/k8s-sidecar:{{ .Values.image.kiwigridTag }}
+          volumeMounts:
+          - name:  api-model
+            mountPath: /config/
+          env:
+          - name: LABEL
+            value: "leapfrogai"
+          - name: FOLDER
+            value: /config/
+          - name: RESOURCE
+            value: both
+          - name: UNIQUE_FILENAMES
+            value: "true"
+          - name: NAMESPACE
+            value: leapfrogai
+        - name: api-container
+          image: ghcr.io/defenseunicorns/leapfrogai-api:{{ .Values.image.lfaiAPITag }}
+          imagePullPolicy: Always
+          env:
+          - name: LFAI_CONFIG_PATH
+            value: /config/
+          - name: LFAI_CONFIG_FILENAME
+            value: "*.toml"
+          - name: PORT
+            value: "{{ .Values.api.port }}"
+          ports:
+            - containerPort: 8080
+          livenessProbe:
+            httpGet:
+              path: /healthz
+              port: 8080
+            initialDelaySeconds: 10
+            periodSeconds: 10
+          readinessProbe:
+            httpGet:
+              path: /healthz
+              port: 8080
+            initialDelaySeconds: 10
+            periodSeconds: 10
+          volumeMounts:
+          - name: api-model
+            mountPath: /config
+      volumes:
+      - name: api-model
+        emptyDir: {}
+
+
+
+

chart/templates/api/permissions.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: read-configmaps
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: Role
+metadata:
+  name: read-configmaps
+rules:
+- apiGroups:
+  - ""
+  resources:
+  - configmaps
+  - secrets
+  verbs:
+  - get
+  - list
+  - watch
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: RoleBinding
+metadata:
+  name: read-configmaps
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: Role
+  name: read-configmaps
+subjects:
+- kind: ServiceAccount
+  name: read-configmaps

chart/templates/api/service.yaml

@@ -0,0 +1,18 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: api
+  annotations:
+    zarf.dev/connect-description: "Load the OpenAPI spec for the LFAI API"
+    zarf.dev/connect-url: "/docs"
+  labels:
+    zarf.dev/connect-name: lfai-api
+spec:
+  selector:
+    app: api
+  ports:
+    - name: http
+      protocol: TCP
+      port: 8080
+      targetPort: 8080
+  type: ClusterIP

chart/templates/namespace.yaml

@@ -0,0 +1,6 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  labels:
+    istio-injection: {{ .Values.istio.injection }}
+  name: leapfrogai

chart/templates/vs.yaml

@@ -0,0 +1,18 @@
+{{- if .Values.istio.enabled }}
+apiVersion: networking.istio.io/v1beta1
+kind: VirtualService
+metadata:
+  name: leapfrogai
+spec:
+  gateways:
+  - istio-system/{{ .Values.istio.gateway }}
+  hosts:
+  - leapfrogai.{{ .Values.domain }}
+  http:
+  - route:
+    - destination:
+        host: api
+        port:
+          number: 8080
+{{- end }}
+---

chart/values.yaml

@@ -0,0 +1,17 @@
+domain: bigbang.dev
+
+image: 
+  lfaiAPITag: 0.1.2
+  kiwigridTag: 1.23.3
+
+istio:
+  enabled: false
+  gateway: tenant
+  injection: disabled
+
+api:
+  replicas: 1
+  port: 8080
+
+monitoring:
+  enabled: false

lfai-values.yaml

@@ -0,0 +1,10 @@
+domain: ###ZARF_VAR_DOMAIN###
+
+image:
+  lfaiAPITag: ###ZARF_CONST_LEAPFROGAI_API_VERSION###
+  kiwigridTag: ###ZARF_CONST_KIWIGRID_VERSION###
+
+istio:
+  enabled: ###ZARF_VAR_ISTIO_ENABLED###
+  gateway: ###ZARF_VAR_ISTIO_GATEWAY###
+  injection: ###ZARF_VAR_ISTIO_INJECTION###

main.py

@@ -4,7 +4,22 @@
 from backends.openai import router as openai_router
 from backends.openai.routes import *
 from utils import get_model_config
+import asyncio
 
 app = FastAPI()
+
+# super simple healthz check
+@app.get("/healthz")
+async def healthz():
+    return {"status": "ok"}
+
+@app.get("/models")
+async def models():
+    return get_model_config()
+
+
+@app.on_event('startup')
+async def watch_for_configs():
+    asyncio.create_task(get_model_config().watch_and_load_configs())
+
 app.include_router(openai_router)
-get_model_config().load()

pyproject.toml

@@ -8,11 +8,15 @@ dependencies = [
     "pyyaml >= 6.0.1",
     "leapfrogai >= 0.3.3",
     "python-multipart >= 0.0.6",
+    "toml >= 0.10.2",
 ]
-requires-python = ">=3.11.4"
+requires-python = ">=3.11.6 <3.12.0"
 
 [project.optional-dependencies]
 dev = ["pip-tools", "pytest", "black", "isort"]
 
 [tool.pip-tools]
 generate-hashes = true
+
+[tool.setuptools.packages.find]
+exclude = ["chart"]