Azure AI Document Intelligence
An Azure service that turns documents into usable data. Previously known as Azure Form Recognizer.
1,978 questions
What could be the reason behind "Contains invalid label name" in Azure Document Intelligence?
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(28.799999999999997, 86%, 12%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EYC%3C/text%3E%3C/svg%3E)
Yan Cahila
0
Reputation points
Hello!
I'm trying to extract data from a scanned document using Azure Document Intelligence and I'm encountering this issue.
The model is successfully trained on the first attempt. Used the trained model to auto-label other documents and upon retraining, this error message is being shown in the logs. And what's weird is it flags the error on a document which was already used on the initial training. What could be the reason on this?
Will really appreciate everybody's insight on this! Thanks in advance.
{count} votes
-
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(6.4, 61%, 10%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ES%3C/text%3E%3C/svg%3E)
santoshkc 13,520 Reputation points Microsoft External Staff2023-12-27T07:27:19.8433333+00:00 Hi @Yan Cahila,
Thank you for reaching out to Microsoft Q&A forum!
I tried to repro your issue, but not able to get the similar error. Iam able to retrain the model on a document that was already used in the initial training. However, you mentioned that you are encountering the error message "Contains invalid label name" when you try to retrain the model on a document that was already used in the initial training.
Try with the different document, to check if similar error persists.
Also I ask you to check the label name used in the initial training is still valid. If the label name is invalid, you can rename it and retrain the model.
Try using Custom extraction model, which gives good accuracy and performance.
I hope you understand! Thank you.
-
santoshkc 13,520 Reputation points Microsoft External Staff
2023-12-28T10:06:59.46+00:00 Hi @Yan Cahila,
Following up to see if the above response was helpful. Thank you!
-
santoshkc 13,520 Reputation points Microsoft External Staff
2023-12-29T07:53:52.46+00:00
Sign in to comment
Sign in to answer