Internal server error after upgrading my document intelligence project to 2024-11-30 API

Question

Internal server error after upgrading my document intelligence project to 2024-11-30 API

Mike Vilters 0

I recently upgraded my project from a preview API (2024-07-31-preview) to the newest general availability API (2024-11-30 API). In this new API version I added signature recognition fields to my custom extraction neural models. However when I now want to analyse a document using my latest trained model, I always get an internal server error. I may have added the signature recognition already in the preview API since it was available, however it didn't work there yet so I upgraded my API and now it should work, but it doesn't.

Is my model now corrupted since i upgraded it from a preview API, or will this resolve in the future? The signature recognition is quite important.

Mike Vilters 0 Reputation points

2025-01-08T14:36:37.3733333+00:00

Even my projects which were initially created in the latest GA API version, receive frequent internal server errors when uploading documents to analyse if there is a signature field present!!
SRILAKSHMI C 18,225 Reputation points Microsoft External Staff Moderator

2025-01-08T19:09:42.75+00:00

Hello Mike Vilters,

Greetings and Welcome to Microsoft Q&A! Thanks for posting the question.

I understand the issue that after upgrading from the 2024-07-31-preview API to the 2024-11-30 API and adding signature recognition fields to a custom extraction neural model, analyzing documents now results in an internal server error,

I tried to repo the scenario it works fine in my environment, here is the screenshot.

Here are some troubleshooting steps:

Verify model compatibility to ensure that the features and configurations used in the model align with the latest API version and its supported functionalities.

Retraining your model using the updated API to avoid potential conflicts or artifacts carried over from previous versions.

Testing the model with a minimal dataset can help isolate the problem by confirming whether the issue lies in the dataset, model configuration, or API usage.

Reviewing logs for error messages and debugging information is crucial for identifying the root cause and resolving any misconfigurations or unexpected behavior effectively.

Also refer this Document Intelligence custom models.

I Hope this helps. Do let me know if you have any further queries.

Thank you!
SRILAKSHMI C 18,225 Reputation points Microsoft External Staff Moderator

2025-01-09T19:54:04.9733333+00:00

Hi Mike Vilters,

Did you get any chance to check the above response. Thank you!
Mike Vilters 0 Reputation points

2025-01-10T08:39:49.4066667+00:00

Yes, I still get these internal server errors. I have retraiend my model, I checked and signatures are supported in the latest version of the API. Now I just created a new project for a new custom model and even there it is still giving me errors. The problem mostly is that the internal server error isn't consistent. Tehre are some models it does work on, and some where it doesn't..
SRILAKSHMI C 18,225 Reputation points Microsoft External Staff Moderator

2025-01-16T22:13:42.5566667+00:00

Hi Mike Vilters,

Please verify that you have the required permissions to access. If the issue persists, kindly share the region you are using and provide a screenshot of the error for further investigation.