Deploy with Helm Chart#

This section shows how to deploy Model Download using Helm chart.

Prerequisites#

Before you begin, ensure that you have the following prerequisites:

Install Helm Chart from Docker Hub or from Source#

To deploy with Helm chart, you can either install the chart from Docker hub or from source.

Option 1: Install from Docker Hub#

  1. Pull the specific chart

    Use the following command to pull the Helm chart from Docker Hub:

    helm pull oci://registry-1.docker.io/intel/model-download-chart --version <version-no>

    See the Docker hub’s tags page for details on the latest version number to use for the application.

  2. Extract the .tgz file

    tar -xvf model-download-chart-<version-no>.tgz

  3. This will create a directory named model-download-chart, containing the chart files. Navigate to the extracted directory:

    cd model-download-chart

Option 2: Install from Source#

  1. Clone the repository containing the Helm chart:

    # Clone the latest on the mainline
  git clone https://github.com/open-edge-platform/edge-ai-libraries.git edge-ai-libraries
# Alternatively, clone a specific release branch
  git clone https://github.com/open-edge-platform/edge-ai-libraries.git edge-ai-libraries -b <release-tag>

  2. Navigate to the chart directory:

    cd edge-ai-libraries/microservices/model-download/chart

Configure the values.yaml File#

Edit the values.yaml file located in the chart directory to set the necessary environment variables. Set your proxy settings as required.

The following is a summary of key configuration options available in the values.yaml file:

Parameter

Description

Example Value

Required

env.HUGGINGFACEHUB_API_TOKEN

Hugging Face access token

hf_xxx

Yes

env.MAX_UPLOAD_SIZE_MB

Maximum allowed upload ZIP size in MB

500

No

env.UPLOAD_CHUNK_SIZE_KB

Chunk size for streaming file uploads in KB (larger improves throughput, smaller reduces memory for concurrent uploads)

8

No

env.EXTERNAL_SOURCES_URL_ALLOWLIST

Comma-separated host/path prefixes allowed for the remote-url hub. When set, it replaces the default allowlist defined in src/plugins/external_sources/sources.yaml

github.com/open-edge-platform/edge-ai-resources/

No

env.GETI_WORKSPACE_ID

Geti™ workspace ID

Yes, For Geti™ connection

env.GETI_HOST

Geti™ connection host address

Yes, For Geti™ connection

env.GETI_TOKEN

Geti™ Personal Access token

Yes, For Geti™ connection

env.GETI_SERVER_API_VERSION

Geti™ API version

v1

Yes, For Geti™ connection

env.GETI_SERVER_SSL_VERIFY

Enables SSL certificate validation for HTTPS/HTTP Geti™ hosts

False

Yes, For Geti™ connection

service.nodePort

Sets the static port (in the 30000–32767 range)

32000

Yes

env.ENABLED_PLUGINS

Comma-separated list of plugins to enable (e.g., huggingface,ollama,ultralytics,pipeline-zoo-models,remote-url,omz,openvino,geti,hls) or all to enable all available plugins

all

Yes

startupConfig.enabled

Creates and mounts a startup-model ConfigMap and schedules its models asynchronously

false

No

startupConfig.config

Startup configuration containing the default download_path, parallel_downloads, and models list

See values.yaml

When enabled

image.repository

image repository url

intel/model-download

Yes

image.tag

latest image tag

latest

Yes

gpu.enabled

For model download deployed on GPU

false

gpu.key

Label assigned to the GPU node on kubernetes cluster by the device plugin example- gpu.intel.com/i915, gpu.intel.com/xe. Identify by running kubectl describe node

<your-node-key-on-cluster>

affinity.enabled

Default is false, true to enable affinity

false

affinity.key

Provide the key for the affinity,default is kubernetes.io/hostname

kubernetes.io/hostname

affinity.value

Provide the values for the respective key

Note: See the chart’s values.yaml file for a full list of configurable parameters.

Preload Models#

Startup model loading is disabled by default. To enable the chart’s ConfigMap-backed, read-only configuration, set:

modeldownload:
  startupConfig:
    enabled: true
    config:
      download_path: /opt/models/preloaded
      parallel_downloads: false
      models:
        - name: BAAI/bge-small-en-v1.5
          hub: huggingface
          type: embeddings
        - name: yolov8n
          hub: ultralytics
          type: vision
          download_path: /opt/models/vision

Enable every plugin referenced by the model entries in modeldownload.env.ENABLED_PLUGINS. The chart sets STARTUP_MODELS_CONFIG to the mounted ConfigMap path. Keep tokens out of startupConfig.config; provide credentials through the existing environment values or your deployment’s secret-injection mechanism.

Model work continues after the readiness probe succeeds. Use the existing jobs endpoints or kubectl logs to monitor jobs. The PVC preserves downloaded artifacts across pod restarts, but in-memory job records are not restored and configured entries are scheduled again. For the full schema and behavior, see Download Models at Startup.

Deploy the Helm Chart#

helm install model-download . -n <your-namespace>

Note: model-download creates and manages a shared PVC that can be used by dependent applications such as Chat Q&A.

Verify the Deployment#

Check the status of the deployed resources to ensure everything is running correctly:

kubectl get pods -n <your-namespace>
kubectl get services -n <your-namespace>

Access the Application#

Open the application’s Swagger documentation in a browser through http://<node-ip>:<node-port>/api/v1/docs.

Uninstall Helm chart#

helm uninstall <name> -n <your-namespace>

Verify the Application#

  1. Ensure that all pods are running and the services are accessible.

  2. Access the application dashboard and verify that it is functioning as expected.

Troubleshooting#

  • If you encounter any issues during the deployment process, check the Kubernetes logs for errors:

    kubectl logs <pod-name>

  • If the PVC created during a Helm chart deployment is not removed or auto-deleted due to a deployment failure or being stuck, delete it manually:

    # List the PVCs present in the given namespace
kubectl get pvc -n <namespace>

# Delete the required PVC from the namespace
kubectl delete pvc <pvc-name> -n <namespace>

Note: Delete the shared PVC only after confirming no other workload or application (for example, Chat Q&A) depends on it. In such cases, uninstall the dependent application first, then clean up model-download resources, and finally delete the shared PVC if required.

Learn More#