Artificial Intelligence Forum
cancel
Showing results for 
Search instead for 
Did you mean: 
Read only

SAP Leonardo Machine Learning Foundation BYOM error

Former Member
0 Likes
1,372

Hi,

I'm still new to SAP Leonardo Machine Learning Foundation and I've encountered an issue when using the BYOM feature.

I have developed an TensorFlow model using version 1.11 which is supported by SAP Leonardo Machine Learning Foundation but when I attempt to create the model server I get the following output:

The model is saved in the required format.

I'm currently using the trial account for SAP Machine Learning Foundation and I was wondering if I can get assistance if anyone has encountered the same issue?

Additionally, does the modelRuntimeId value represent the version of TensorFlow?

1 ACCEPTED SOLUTION
Read only

Former Member
0 Likes
1,235

These are the parameters I used in creating the correct runtime environment using the DEPLOYMENT_API_URL provided in my service-key.

I entered the above url in the browser and to load the Deployment API using swagger.

Click on model-server-controller->Creates a model server

In the parameters section enter the following:

Parameter: modelServerSpecsRequest

Value: { "specs": { "enableHttpEndpoint": false, "modelRuntimeId": "%Your-TensorFlow-version%", "models": [ { "modelName": "%Your-Model-Name%", "modelVersion": "%Your-Model-Version%" } ], "replicas": 1, "resourcePlanId": "%Your-Resource-Plan%" } }

Parameter content type: application/json
Enter your authorization and then try out the request.
I hope this helps for anyone who has a such an issue.

View solution in original post

Hi,

I'm still new to SAP Leonardo Machine Learning Foundation and I've encountered an issue when using the BYOM feature.

I have developed an TensorFlow model using version 1.11 which is supported by SAP Leonardo Machine Learning Foundation but when I attempt to create the model server I get the following output:

The model is saved in the required format.

I'm currently using the trial account for SAP Machine Learning Foundation and I was wondering if I can get assistance if anyone has encountered the same issue?

Additionally, does the modelRuntimeId value represent the version of TensorFlow?

1 REPLY 1
Read only

Former Member
0 Likes
1,236

These are the parameters I used in creating the correct runtime environment using the DEPLOYMENT_API_URL provided in my service-key.

I entered the above url in the browser and to load the Deployment API using swagger.

Click on model-server-controller->Creates a model server

In the parameters section enter the following:

Parameter: modelServerSpecsRequest

Value: { "specs": { "enableHttpEndpoint": false, "modelRuntimeId": "%Your-TensorFlow-version%", "models": [ { "modelName": "%Your-Model-Name%", "modelVersion": "%Your-Model-Version%" } ], "replicas": 1, "resourcePlanId": "%Your-Resource-Plan%" } }

Parameter content type: application/json
Enter your authorization and then try out the request.
I hope this helps for anyone who has a such an issue.