Snowflake Model Registry user interface¶
Note
Model Registry Snowsight UI is generally available in all deployments.
Inference Services UI for SPCS Model Serving is in preview in AWS, Azure and GCP commercial deployments only.
On the Models page in Snowsight, you can find all your machine learning models. You can also view their metadata and deployments.
Model details¶
The Models page displays the models that you’ve created and logged into the Snowflake Model Registry or have access to the models via the USAGE or READ privilege. These are both models that have been developed with Snowpark ML and externally sourced models (such as models from Hugging Face). It also shows Cortex Fine-tuned models, and may include other model types that you can create in Snowflake in future releases.
To display the Models page, in the navigation menu, select AI & ML » Models. The resulting list contains all the models in the Snowflake Model Registry in all the databases and schemas that your current role has access to.
Note
If you don’t see any models, make sure your role has the required privileges.
To open a model’s details page, select the corresponding row in the Models list. The details page displays key model information, including the model’s description, tags, and versions.
To edit the model description or delete the model, select … in the top right corner.
To open the version’s details page, select a model version. This page displays model version metadata, such as metrics, and a list of available methods that can be called from Python or SQL.
To view code that calls the model function, select the SQL or Python link next to it. You can copy this code snippet into a Snowsight SQL worksheet or a Python notebook.
To add or modify metadata or delete the model version, select the … in top right corner.
The Files tab contains a list of the model version’s underlying artifacts. You can download individual files from this page. This page is only available if the user has either OWNERSHIP or READ privilege on the model.
The Lineage tab shows the full data flow lineage information for the model, including any datasets that were used to train the model, any feature views from Feature Store, and the source data tables.
Deploy user models¶
You can deploy models to SPCS Model Serving directly from the Model Registry page.
Note
Snowflake Model Registry only supports deploying user models to SPCS Model Serving.
To deploy a model, complete the following steps:
Select a model from the list of models.
From the model details page, navigate to the Versions section.
To open the version details page, select a model version from the list of versions.
From the version details page, select the Deploy button.
From the opened pane, enter a name for the service to be deployed.
Select whether to create a REST API endpoint for the deployed service.
Select a compute pool for the deployed service.
(Optional) To customize performance and resource usage, you can adjust details, such as the number of workers, CPU, and memory, from the advanced settings.
Select Deploy.
The deployment process can take up to 15 minutes to create the service.
After the deployment is complete, the you can view the service from the Inference Services tab on the main Model Registry page.
Model inference services¶
You can see the model inference services created with SPCS Model Serving in the Model Registry UI. The main model listing page shows the status of inference services created for any model.
If you select model name and a model version, you can use the Inference Services tab in the model version details page to see more details about the deployed inference service, as well as suspend the inference service. This also shows the list of functions that the service exposes. And you can see or copy the SQL or Python usage code snippet.
Select Open Details to display service parameters. To view more details about the deployed inference service, select Open Service Details from the service parameters pane. You can also access the service details from the Inference Services tab on the main Model Registry page.
Model monitoring¶
For any models that have Model Monitors attached to them, you can visualize model monitoring metrics using the Model Monitors in the model details page.
Select the desired model monitors to display the Monitoring dashboard:
Select Compare to view the menu of model version select a second model version to compare this model version against:
Monitoring supports a large number of model accuracy, model drift, and feature drift metrics. To select the metrics that are computed and displayed, select Settings icon to choose the desired metrics:
