ModelBuildError: Could not build the model. 'Contains undefined table label name.'

Question

ModelBuildError: Could not build the model. 'Contains undefined table label name.'

Martynas Trusilo 0

Hi,

I am training a custom extraction model. I have labeled all the data that I NEED. Filler sentences are obviously left out as that is useless information. I am trying to train model, however, I keep getting the folowing error:

ModelBuildError: Could not build the model: Can't find any valid labels for provided dataset. Labels file filename.pdf is invalid for the following reason(s): 'Contains undefined table label name.'

What am I doing wrong - I have done everything exactly as per video provided on the introductory of custom extraction models.
I have checked every single label and all are labeled and contain information. Every cell of the table that I used does contain information. What is causing this?

Thanks!

Martynas Trusilo 0 Reputation points

2024-08-27T10:23:30.95+00:00

I have already done that, what kind of different approach are you talking about?
santoshkc 15,325 Reputation points Microsoft External Staff Moderator

2024-08-28T16:52:47.26+00:00

Hi @Martynas Trusilo,

It seems the error is due to inconsistencies in your label definitions or table structures. Double-check that all labels in your dataset match the schema defined in your project without any typos, extra spaces, or missing labels. Sometimes, empty or invisible labels can cause such errors, so thoroughly check for any fields that might appear labeled but lack proper values. If the issue persists, try training with a smaller, simplified dataset to pinpoint the problematic labels. Rebuilding the label file or re-annotating a portion of your dataset can also help resolve any hidden inconsistencies.

See: Build and train a custom extraction model.

I hope you understand! Thank you.
santoshkc 15,325 Reputation points Microsoft External Staff Moderator

2024-08-29T17:37:20.12+00:00

Hi @Martynas Trusilo,

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Thank you.

1 answer

Your answer

Martynas Trusilo 0 Reputation points

2024-08-27T10:23:30.95+00:00

I have already done that, what kind of different approach are you talking about?
santoshkc 15,325 Reputation points Microsoft External Staff Moderator

2024-08-28T16:52:47.26+00:00

Hi @Martynas Trusilo,

It seems the error is due to inconsistencies in your label definitions or table structures. Double-check that all labels in your dataset match the schema defined in your project without any typos, extra spaces, or missing labels. Sometimes, empty or invisible labels can cause such errors, so thoroughly check for any fields that might appear labeled but lack proper values. If the issue persists, try training with a smaller, simplified dataset to pinpoint the problematic labels. Rebuilding the label file or re-annotating a portion of your dataset can also help resolve any hidden inconsistencies.

See: Build and train a custom extraction model.

I hope you understand! Thank you.
santoshkc 15,325 Reputation points Microsoft External Staff Moderator

2024-08-29T17:37:20.12+00:00

Hi @Martynas Trusilo,

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Thank you.

Answer 1

Hi @Martynas Trusilo,

Thank you for reaching out to Microsoft Q&A forum!

The error message suggests that there’s an issue with your label file, which indicates that one of the labels in your file isn’t recognized by the model.

To fix this, review your label file to ensure that all labels are correctly defined. Make sure that the label names match those used in your training data, with no typos or inconsistencies. Additionally, confirm that the labels are applied consistently throughout the file.

If the problem persists, consider using a different label file or consult the documentation for your custom extraction model to check for any specific label file requirements or limitations.

Please look into: Tips for building labeled datasets.

I hope you understand! Thank you.

Share via

ModelBuildError: Could not build the model. 'Contains undefined table label name.'

1 answer

Your answer