dlim

Edit online

Usage

dlim --help Displays help information.

dlim --version Version information.

dlim config configuration-options

dlim model model-options

dlim runtime runtime-options

Commands

--help Displays help information.

--version Version information.

config Configuration information to access inference services.

model Manage models for inference services.

runtime Manage runtime for inference services.

Model options

list List all models deployed for an inference service.

view View details of a model deployed on an inference service.

start Start the services of a model from an inference service.

stop Stop the services of a model from an inference service.

undeploy Undeploy a model from an inference service.

viewprofile View the profile of a model deployed on an inference service.

updateprofile Update the profile of a model deployed on an inference service.

Runtime options

list List all available runtimes for an inference service.

view View details of a runtime for an inference service.

create Create a new runtime into inference service.

delete Delete a runtime from an inference service.

update Update the profile of a model deployed on an inference service.

Examples

Model options:

List all models:
```
dlim model list
```
View model details:
```
dlim model view model-name
```
Start a model:
```
dlim model start model-name
```
Stop a model:
```
dlim model stop model-name
```
Deploy the model from a specified location. For example deploy the model in folder folder-name
```
dlim model deploy -p folder-name
```
Deploy the model from a specified location. For example deploy the model from archived file archived-file:
```
dlim model deploy -p archived-file
```
Undeploy the model:
```
dlim model undeploy model-name
```
View a model profile:
```
dlim model viewprofile model-name
```

Update a model profile:

dlim model updateprofile model-name -f profile.json

Runtime options:

List all runtimes:
```
dlim runtime list
```
View runtime details:
```
dlim runtime view runtime-name
```
Create a runtime:
```
dlim runtime create -f runtime.json
```
Delete a runtime:
```
dlim runtime delete runtime-name
```

Update a runtime:

dlim runtime update runtime-name -f runtime.json

Note

The dlim command stores its configuration data in the user home directory. If you are running dlim inside a pod or an alternate environment, the user home directory is not writable. To avoid any permission errors trying to access the user home directory, use a JSON web token (JWT) token to bypass configuration caching in the user home directory. For example, in 2.3.6 and later, use:

$dlim model list --rest-server <EDI_URL> --jwt-token <USER_ACCESS_TOKEN>

For example:

$dlim model list --rest-server https://wmla-console-wmla.apps.myocp.example.com/dlim/v1/ --jwt-token $jwt_token

The JSON web token (JWT) is generated by the platform. To obtain an authorization token, refer to the Cloud Pak for Data Platform API or Generating an authorization token. If you are using Watson Studio runtimes, the JWT token can be found in the environment variable USER_ACCESS_TOKEN or DSX_TOKEN. For example:

$dlim model list --rest-server https://wmla-console-wmla.apps.myocp.example.com/dlim/v1/ --jwt-token $USER_ACCESS_TOKEN