Skip to main content

Batch — Models

This article refers to Platform v3.1.0. The current Platform version is v3.2.0.

Overview​

The Models category of the Batch Command Manager lets you operate AI models across many nodes at once. The four available commands are Install, Delete, Run, and Stop.

Models options in the batch wizard

Models options in the batch wizard

Commands​

1. Install​

Deploy a model and version from the App Library to every selected node. The form exposes three advanced toggles:

  • Enable GPU usage — schedule the model on the node's GPU when available.
  • Activate logs — start capturing logs immediately after install.
  • Start serving automatically after install — bring the model online without an extra Run step.
Install policy for models

Install policy for models

note

If the model is already installed on a target node, pick the desired behaviour:

  • Install a new instance of the model.
  • Overwrite the already installed model.
  • Update — overwrite only if the installed version is outdated.
  • Do not install the model.

2. Delete​

Remove the selected model from every target node.

note

If a node has more than one instance of the model, the Delete command is applied to every instance in parallel.

3. Run​

Start the selected model on every target node where it is installed but stopped.

note

If a node has more than one instance of the model, the Run command is applied to every instance in parallel.

4. Stop​

Stop the selected model on every target node where it is currently serving.

note

If a node has more than one instance of the model, the Stop command is applied to every instance in parallel.

Summary​

The Models batch category is the bulk parallel of the single-node model deployment flow described in Models. Use it when the same model needs to roll out — or be retired — across a whole fleet of edge nodes.