Viewing ModelArts Model Details

Viewing the Model List

You can view all created models on the model list page. The model list page displays the following information.

**Table 1** Model list
Parameter	Description
Model Name	Model name.
Latest Version	Latest version of a model.
Status	Model status.
Deployment Type	Types of the services that a model can be deployed as.
Versions	Number of model versions.
Request Mode	Request mode of real-time services. Synchronous request: one-off inference with results returned synchronously (within 60s). This mode is suitable for images and small video files. Asynchronous request: one-off inference with results returned asynchronously (longer than 60s). This mode is suitable for real-time video inference and large videos.
Created	Model creation time.
Description	Model description.
Operation	Deploy: Deploy a model as real-time services, edge services, or batch services. Create Version: Create a model version. The settings of the last version are used by default, except for the version. You can change the parameter settings. Delete: Delete a model. NOTE: If a model version has been deployed as a service, you must delete the associated service before deleting the model version. A deleted model cannot be recovered.

Click the number in Versions to view the version list.

Figure 1 Version list

The version list displays the following information.

**Table 2** Version list
Parameter	Description
Version	Current version of a model.
Status	Model status.
Deployment Type	Types of the services that a model can be deployed as.
Model Size	Model size.
Model Source	Model source.
Created	Model creation time.
Description	Model description.
Operation	Deploy: Deploy a model as real-time services, edge services, or batch services. Delete: Delete a version of a model.

Viewing Model Details

After a model is created, you can view the model information on the model details page.

Click the name of the target model to access its details page.

On the model details page, you can view the basic information and precision of the model, and switch tab pages to view more information.

**Table 3** Basic model information
Parameter	Description
Name	Model name.
Status	Model status.
Version	Current version of a model.
ID	Model ID.
Description	Click the edit button to add the description of a model.
Deployment Type	Types of the services that a model can be deployed as.
Meta Model Source	Source of the meta model, which can be training jobs, OBS, or container images.
Training Name	Associated training job if the meta model comes from a training job. Click the training job name to go to its details page.
Training Version	Training job version if the meta model comes from an old-version training job.
Storage path of the meta model	Path to the meta model if the meta model comes from OBS.
Container Image Storage Path	Path to the container image if the meta model comes from a container image.
AI Engine	AI engine if the meta model comes from a training job or OBS.
Engine Package Address	Engine package address if the meta model comes from OBS and AI Engine is Custom.
Runtime Environment	Runtime environment on which the meta model depends if the meta model comes from a training job or OBS and a preset AI engine is used.
Container API	Protocol and port number for starting the model if the meta model comes from OBS (AI Engine is Custom) or a container image.
Inference Code	Path to the inference code if the meta model comes from an olde-version training job.
Image Replication	Image replication status for meta models from a container image.
Dynamic loading	Dynamic loading status if the meta model comes from a training job or OBS.
Size	Model size.
Health Check	Displays health check status if the meta model comes from OBS or a container image. When health check is enabled, the probe parameter settings are displayed. Startup Probe: This probe checks if the application instance has started. If a startup probe is provided, all other probes are disabled until it succeeds. If the startup probe fails, the instance is restarted. If no startup probe is provided, the default status is Success. Readiness Probe: This probe verifies whether the application instance is ready to handle traffic. If the readiness probe fails (meaning the instance is not ready), the instance is taken out of the service load balancing pool. Traffic will not be routed to the instance until the probe succeeds. Liveness Probe: This probe monitors the application health status. If the liveness probe fails (indicating the application is unhealthy), the instance is automatically restarted. The probe parameters include Check Mode, Health Check URL (displayed when Check Mode is set to HTTP request), Health Check Command (displayed when Check Mode is set to Command), Health Check Period, Delay, Timeout, and Maximum Failures.
Model Description	Description document added during the creation of a model.
Instruction Set Architecture	System architecture.
Inference Accelerator	Type of inference accelerator cards.

**Table 4** Model details tabs
Parameter	Description
Model Precision	Model recall, precision, accuracy, and F1 score of a model.
Parameter Configuration	API configuration, input parameters, and output parameters of a model.
Runtime Dependency	Model dependency on the environment. If creating a job failed, edit the runtime dependency. After the modification is saved, the system will automatically use the original image to create the job again.
Events	The progress of key operations during model creation. Events are stored for three months and will be automatically cleared then. For details about how to view events of a model, see Viewing ModelArts Model Events.
Constraint	Displays the constraints of service deployment, such as the request mode, boot command, and model encryption, based on the settings during model creation. For models in asynchronous request mode, parameters including the input mode, output mode, service startup parameters, and job configuration parameters can be displayed.
Associated Services	The list of services that a model was deployed. Click a service name to go to the service details page.