Manage Models

Rename, update keys, and delete trained models in Datature Vi to keep your deployments organized.

Datature Vi gives you three operations for managing trained models: rename, edit model key, and delete. All three are accessed from the Models tab inside any training project.

This section covers each operation so you know when to use it and what to expect.

Models are referenced internally by their run ID, not by their display name or model key. Renaming a model or updating its key does not affect active inference or SDK queries.

By the end of this guide

Rename, edit model keys, and delete trained models from your deployments.


Model management operations

Rename A Model

Update the display name to reflect the model's purpose, version, or deployment stage.

Edit A Model Key

Update the checkpoint identifier to a human-readable name for better organization.

Delete A Model

Permanently remove model weights, adapters, and configuration. This action cannot be undone.


How to access model operations

All model management options are available from the same menu:

1

Open your training project

Open your training project

Go to Training in the sidebar and click the project containing the model you want to download.

The model card menu lists six actions. Each one opens a different part of the product or starts a workflow.

Model card menu options

Option
What it does
View Run
Opens the training run that produced this model so you can review metrics, logs, and run settings.
Export Model
Opens the Export Model dialog so you can generate a pre-filled Vi SDK Python snippet. Run that code on your machine to pull model files into your local cache.
Create Deployment
Reserved for hosted deployment from the platform. This option is not available yet.
Rename Model
Changes the model's display name in the UI and in API responses. See Rename a model.
Edit Model Key
Changes the checkpoint identifier used to organize and reference the model. See Edit a model key.
Delete Model
Permanently removes the model's weights, adapters, and stored configuration. See Delete a model.

Before You Delete

Deleting a model is permanent. Download the model first if you may need it later. See Download a model for the snippet workflow and SDK download steps.


Next steps

Download A Model

Generate a Vi SDK snippet, then run it locally to fetch weights and configuration.

Manage Runs

Review training history, metrics, and checkpoints from completed runs.

Vi SDK Reference

Programmatic access to model listing, downloading, and renaming.