Image Management

Overview

During the development and runtime of AI services, complex environment dependencies need to be debugged for containerization. In the best practices of AI development in ModelArts, container images are used to provide fixed runtime environments. In this way, dependencies can be managed and the runtime environments can be easily switched. The container resources provided by ModelArts enable quick and efficient AI development and model experiment iteration.

The preset images provided by ModelArts by default have the following features:

• Out-of-the-box and scenario-specific: Typical dependent environments for AI development are preset in these images to provide optimal software, OS, and network configurations. They have been fully tested on hardware to ensure optimal compatibility and performance.
• Configuration customizable: Preset images are stored in the SWR repository for you to customize and register them as your own images.
• Secure and reliable: Access policies, user permissions control, vulnerability scanning for development software, and OS are configured based on best practices for security hardening to ensure the security of images.

If you have special requirements on the deep learning engine and development library, you can use ModelArts custom images to customize runtime engines.

Based on the container technology, you can customize container images and run them on ModelArts. Custom images support CLI parameters and environment variables in free text format, featuring high flexibility for a wide range of compute engines.

Application Scenarios of Preset Images

ModelArts provides a group of preset images. You can use a preset image to create a notebook instance. After installing and configuring dependencies on the instance, create a custom image. Then, you can directly use the image in ModelArts for training jobs without any adaptation. You can also use preset images to submit training jobs and create AI applications.

We recommend the preset image version based on your development requirements and stability of the version. If your development can be carried out using versions preset in ModelArts, for example, MindSpore 1.X, use the preset images. They have been fully verified and have many commonly-used installation packages, relieving you from configuring the environment.

Application Scenarios of Custom Images

• Using custom images on notebook instances

  If the preset images of notebook instances cannot meet requirements, you can create a custom image by installing and configuring the software and other data required by the environment in a preset image. Then, use the custom image to create new notebook instances.

• Using a custom image to create training jobs

  If you have developed a model or training script locally but the AI engine you used is not supported by ModelArts, create a custom image and upload it to SWR. Then, use this image to create a training job on ModelArts and use the resources provided by ModelArts to train models.

• Using a custom image to create AI applications

  If you have developed a model using an AI engine that is not supported by ModelArts, to use this model to create AI applications, do as follows: Create a custom image, import the image to ModelArts, and use it to create AI applications. The AI applications created in this way can be centrally managed and deployed as services.

Custom Image Services

When you use a custom image, the following services may be involved:

• SWR

  Software Repository for Container (SWR) provides easy, secure, and reliable management over container images throughout their lifecycle, facilitating the deployment of containerized applications. You can upload, download, and manage container images through the SWR console, SWR APIs, or community CLI.

  Your custom images must be uploaded to SWR. The custom images used by ModelArts for training or creating AI applications are obtained from the SWR service management list.

  Figure 1 Obtaining images
  
• OBS

  Object Storage Service (OBS) is a cloud storage service optimized for storing massive amounts of data. It provides unlimited, secure, and highly reliable storage capabilities at a relatively low cost.

  ModelArts exchanges data with OBS. You can store data in OBS.

• ECS

  An Elastic Cloud Server (ECS) is a basic computing unit that consists of vCPUs, memory, OS, and Elastic Volume Service (EVS) disks. After an ECS is created, you can use it similarly to how you would use your local PC or physical server.

  You can create a custom image on premises or on an ECS.

When you use a custom image, ModelArts may need to access dependent services, such as SWR and OBS. The custom image can be used only after the access is authorized. It is a good practice to use an agency for authorization. After the agency is configured, the permissions to access dependent services are delegated to ModelArts so that ModelArts can use the dependent services and perform operations on resources on your behalf. For details, see Configuring Access Authorization (Global Configuration).