Azure Form Recognizer documentClassifiers training

FJackson 35 Reputation points
2023-06-06T21:10:41.05+00:00

Hi ,

I want to train my Azure Form Recognizer custom classification model. According to the REST API documentation. However building the model with the following code fails with the following error message:

"TrainingContentMissing: Training data is missing: Could not find any training data at the given path."

The training source is uploaded in the Azure Blob Storage ad the right prefix folders. If I use the Form Recognizer Studio, it finds the training source and asks me if it should label it according to the folder name.

https://westeurope.api.cognitive.microsoft.com/formrecognizer/documentClassifiers:build?api-version=2023-02-28-preview


Body:
{
  "classifierId": "customClassificationModelVersion4",
  "description": "",
  "docTypes": { 
 "Label1": {
      "azureBlobSource": {
containerUrl": "[CONTAINER SAS TOKEN, inluding the containername]",
        "prefix": "training/Label1/"
      }
    }, 
 "Label2": {
      "azureBlobSource": {
        "containerUrl": "[CONTAINER SAS TOKEN, inluding the containername]",
        "prefix": "training/Labe2/"
      }
    }, 
 "Label3": {
      "azureBlobSource": {
        "containerUrl": "[CONTAINER SAS TOKEN, inluding the containername]",
        "prefix": "training/Label3/"
      }
    }, 
 "Label4": {
      "azureBlobSource": {
        "containerUrl": "[CONTAINER SAS TOKEN, inluding the containername]",
        "prefix": "training/Label4/"
      }
    }
  }
}
Azure AI Document Intelligence
{count} votes

2 answers

Sort by: Most helpful
  1. Ahmet Arsan 0 Reputation points Microsoft Employee
    2024-11-06T18:41:35.81+00:00

    im also having this issue.


  2. Dyaneshwaran Sivashanmugam 21 Reputation points
    2025-12-02T22:47:54.8+00:00

    is this issue resolved ? i get same error now . error message is not giving any inputs on where exactly error is occurring

    0 comments No comments

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.