From PVC (UI)

This guide explains how to add a model that’s already been pre-loaded on a Persistent Volume Claim (PVC) within your Kubernetes namespace for MLIS.

Before You Start #

Create a PVC and pre-load a model
Obtain the model’s path within the PVC
Review model-specific resources (e.g., NVIDIA NIM documentation) that you may need for adding specific environment variables and arguments during packaged model creation.

Review the following PVC syntax and URL options to ensure you have the correct information for adding your model.

Option	Description	Example	Default
PVC Name	Name of the Persistent Volume Claim (PVC) to be mounted	`pvc://my-model-pvc`	Required, no default
Path	Optional path within the PVC to be mounted	`pvc://my-model-pvc/models`	If not specified, the entire PVC is mounted
ContainerPath	Directory in container where the PVC is mounted	`pvc://my-model-pvc?containerPath=/mnt/models`	`/mnt/models`
readOnly	Whether the volume is read-only	`pvc://my-model-pvc?readOnly`	If not specified, the volume is read-write

PVC Name

The PVC must exist in the Kubernetes namespace where the packaged model will be deployed.

Sign in to HPE Machine Learning Inferencing Software and navigate to Packaged Models.
Select Add new model and fill in the basic details.
In Storage Details:
- Set Registry to None and Model Format to Custom.
- Enter the appropriate Image name.
- Specify the URL/Path using PVC syntax (e.g., pvc://models-cache-pvc?containerPath=/mnt/models).

Choose a Resource Template or define custom resources.

Name	Description	Request CPU	Request Memory	Request GPU	Limit CPU	Limit Memory	Limit GPU
cpu-tiny	1 cpu, 10Gi memory, no gpu per replica	1	10Gi		1	10Gi
cpu-small	4 cpu, 20Gi memory, no gpu per replica	4	20Gi		6	40Gi
cpu-large	8 cpu, 40Gi memory, no gpu per replica	8	40Gi		10	60Gi
gpu-tiny	1 cpu, 10Gi, 1 gpu per replica	1	10Gi	1	1	10Gi	1
gpu-small	2 cpu, 20Gi, 2 gpu per replica	2	20Gi	2	6	40Gi	2
gpu-large	8 cpu, 40Gi, 4 gpu per replica	8	40Gi	4	10	60Gi	4

GPU Type

Specifying a GPU type requires heterogenous GPU support be enabled.

Set any necessary Environment Variables and Arguments based on your packaged model’s framework type. For more information, see the Advanced Configuration reference article.
Select Create model.