Manage a Model Registry

Manage a Model Registry

The Model Registry serves as a centralized repository to track, organize, and manage your AI/ML models. It ensures reproducibility, versioning, and simplifies deployment. Follow this guide to understand how to manage your model registry and prepare configurations needed for deploying a model endpoint.

Accessing the Model Registry

  • Log in to your GreenNode AI Platform account and navigate to the Model Registry Dashboard.
  • Find and click on the "Import a model registry" button.

Import a Model Registry

  • Location & Model registry name: Choose the target location and assign a unique name for your model in the registry.
  • Container: Select one of the following options:
    • Pre-built Container: Use a supported framework maintained by the platform.
    • Custom Container: Provide your own Docker image for full control.
  • Pre-built Container Configuration (if selected)
    • Framework: Choose a model deployment framework & suitable version that meets your requirements (e.g., SGLang 0.4.3, vLLM 0.8.5, etc.)
    • Model source: Access to model stored: from our network volume, from the GreenNode catalog, or directly from Hugging Face.
      Info
      • If select from HuggingFace, it is recommended to select a network volume. Choose a data source where the model is already cached to speed up loading. This ensures the model is cached, reducing loading time for future use.
    • Framework Settings: Configure parameters for the selected framework
  1. Custom Container Configuration (if selected)
    1. Image URI: Provide the full URI of your Docker image (e.g., from Docker Hub or a private registry).
    2. Username and Password: Provide credentials if your image requires basic authentication.
    3. Command & Args (optional): Override the default command or entrypoint if needed for model serving.
    4. Access Port: Specify the port on which the model service is exposed inside the container.
  2. Click the “Import” button to complete the process.

Edit a Model Registry

You can update metadata or configuration settings for an existing model in the registry, provided it is not currently deployed or active.
  1. Navigate to the Model Registry dashboard.  
  2. Locate the model entry you want to update.  
  3. Click the "Edit" icon next to the model.  
  4. Modify fields such as:  
    1. Model name  
    2. Framework
    3. Framework settings (if applicable)  
  5. Click "Save Changes" to apply updates.
Notes
Note: If the model registry is active (i.e., deployed to an endpoint), editing is not allowed. You must undeploy or clone the model to make changes.
Clone a Model Registry
Cloning allows you to create a new model registry entry with the same configuration, artifacts, and framework settings. This is useful for versioning, testing, or deploying in a different environment.
  1. In the Model Registry, find the model you want to clone.  
  2. Click the "Clone" button.  
  3. Enter a new model name for the clone.  
  4. (Optional) Modify description, tags, or minor metadata.  
  5. Click "Clone" to create the new registry entry.  
Info
The cloned model will be independent of the original and can be edited or deployed separately.

Delete a Model Registry

Deleting a model registry entry will permanently remove the model, all its versions, and related artifacts from the platform.
  1. Locate the model you wish to delete in the Model Registry.  
  2. Click the "Delete" icon next to the model entry.  
  3. Confirm the deletion in the pop-up dialog.

    • Related Articles

    • Manage a model endpoint

      This guide will walk you through the key features and steps involved in deploying your models, optimizing costs through undeployment, and removing endpoints when they are no longer needed. After creating a model endpoint, follow these steps to ...

    • Import a model registry with custom container

      The model registry is a centralized repository for storing trained models, their metadata, versions, and associated artifacts. It allows for version control, management, and organization of models developed during the training phase. This enables ...

    • Import a Model Registry using Triton Server

      Model Preparation Since our AI Platform only accesses models from a Network Volume, you must first create a Network Volume. Pull your model from local file systems or cloud storage (AWS S3, Azure Blob, or GCS) into the Network Volume. Ensure the ...

    • Import a Model Registry using NVIDIA NIM

      Step 1: Accessing the Model Registry Log in to your GreenNode AI Platform account and navigate to the Model Registry Dashboard Model Registry Dashboard at: . Find and click on the "Import a model registry" button. Step 2: Import a Model Registry ...

    • Manage a model tuning job

      Model tuning, also known as hyperparameter optimization, is the process of adjusting the hyperparameters of a machine learning model to improve its performance. Hyperparameters are settings that determine the learning process of a model and are not ...