Application Scenarios

ExeML projects cover data labeling, model training, and deployment. No charges will be incurred for data labeling. Compute resources are required for model training and deployment. You are charged based on the resources you select. Currently, ModelArts provides free compute flavors for a limited time, allowing you to experience ExeML free of charge.

Figure 1 Free flavors provided by ExeML

The free resources for ExeML training do not include the OBS storage resource. Data stored in OBS is billed based on the OBS specifications you use.

Declaration for Free Flavors (Model Training)

• The free ExeML training flavor is used for trials. The training job will automatically stop in one hour (excluding the time required for publishing the model). Therefore, you are advised to set the maximum training duration to one hour.
• The free flavors with limited performance are recommended for model training if your data volume is large or the training duration exceeds one hour.
• The free flavors are limited. If there are a large number of users, you must wait in a queue for a long time. For better experience, you are advised to select a pay-per-use flavor.
• A free flavor can be used in only one training job. After the training job is complete, other training jobs can use the free flavor again.
• Free flavors are available only in the CN North-Beijing4 region.

Declaration for Free Flavors (Deployment)

• The free flavors are used only for trials. The deployed service will automatically stop in one hour (excluding the time required for publishing the model). If you want to continue running the service using the free flavor, restart the service. You can run the service for another hour. After one hour, the service automatically stops.
• The free flavors are limited. If there are a large number of users, you must wait in a queue for a long time. For better experience, you are advised to select a pay-per-use flavor.
• Free (CPU) or Free (GPU) are free of charge for service deployment.
• A free flavor can be used in only one service. If a free flavor has been used for deployment in an ExeML project, no matter whether the service is running, other deployment tasks cannot use this free flavor.
• Free flavors are available only in the CN North-Beijing4 region.

Trying ExeML for Free

The following uses image classification as an example. The procedure for other types of ExeML projects is similar. Before performing the following operations, complete account registration and global configuration by referring to Preparations. This document describes how to use ExeML free of charge. For details about how to use ExeML, see the Getting Started.

1. Log in to the ModelArts management console and choose ExeML from the left navigation pane.
2. Prepare a dataset for model training by referring to Preparing Data and upload the dataset to OBS as required.
3. Create an ExeML project for Image Classification by referring to Creating a Project. In the project, select the data prepared in 2.
4. Label all image data by referring to Labeling Data.
5. In the Training Configuration dialog box, select Free (GPU) for Instance Flavor, select I have read and agree to the above, and click Yes to start model training. For details, see Performing Auto Training.
   Figure 2 Selecting the free flavor
   
6. On the Train Model tab page, wait until training is complete. When the training status changes to Completed, you can view the training details on the current page.

   If you are satisfied with the training result, deploy the model as a service. Otherwise, you are advised to optimize the data and labeling information and then start the training again.

7. Deploy the trained model as a real-time service. For details, see Deploying a Service.

   In the Deploy dialog box that is displayed, you can select Free (CPU) or Free (GPU). Both flavors are free of charge. The number of compute nodes must be set to 1. Select I have read and agree to the above and click OK to start the deployment task.

   Figure 3 Selecting a free flavor to deploy a real-time service
   
8. When the status of the deployment task changes to Running, the deployment is complete. On the Deploy Service tab page, you can view the URL of the service. Alternatively, you can upload an image on this page for predictive recognition.

   The service that uses the free flavor will automatically stop in one hour. If you need to use the service, click Start in the Version Manager pane. The system starts the service using the selected free flavor. If the free flavor is being used by other services of the current account, the service needs to wait in a queue.