irob772 avatar image
1 Vote"
irob772 asked irob772 commented

Health Text Analytics container throws ModelException

I have retrieved the latest Health Text Analytics container, and when I run it locally it will load, but any request produces the following exception (full output is attached):

fail: Microsoft.CloudAI.Containers.TextAnalytics.Health.Processor.HealthV3Dot1Preview4Processor[0]
Microsoft.CloudAI.Containers.TextAnalytics.TATK.Exceptions.ModelException: The NER model is currently loading. Please try again soon. at Microsoft.CloudAI.Containers.TextAnalytics.HealthcareClient.HealthcareClient.BatchPredictFromGrpcAsync(IList`1 batchInput, Boolean isExternalRequest, RequestCorrelation correlation, String modelVersion, CancellationToken cancellationToken) in /source/src/container/src/Health/Client/HealthcareClient.cs:line 234

When calling /ready the result is always "unready". I was able to deploy the container to an Azure Docker Web App successfully, so I believe my end point/api key configuration is correct. This has happened on two computers so far. Is this a resource issue?

Thanks for the help!

· 2
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.


Is there any document you are referring to? Could you please share it so that we can try to reproduce this issue? Thanks.


0 Votes 0 ·
irob772 avatar image irob772 YutongTie-MSFT ·

Hi Yutong,

I'm working from the documentation at

I can download and run the container using the scripts there as mentioned, but get this exception when attempting to execute the /ready endpoint or attempt to process a document in the /demo area.

The script to run the container is below:

 docker run --rm -it -p 5000:5000 --cpus 6 --memory 12g \ \
 Eula=accept \
 rai_terms=accept \
 Billing={ENDPOINT_URI} \
 ApiKey={API_KEY} \

In order to get the logging output in the console, remove the Logging:Disk:Format=json parameter


0 Votes 0 ·

0 Answers