Marketplace user guide

Overview​

• Train Model
• Get Model Info
• Get Model Keys
• Delete Model

1. Establishes a connection to your Azure Form Recognizer resource (Azure Form Recognizer Scope).

2. Creates a new custom training model using the source form documents that you stored during Setup (Train Model).

3. Retrieves and outputs the status of the train model operation (Do While, Get Model Info, and Write Line).

4. Retrieves and outputs the keys extracted during the training model operation (Get Model Keys, For Each, and Write Line).

5. Deletes the new custom training model that you created (Delete Model).

   

Prerequisites​

1. Complete the Azure Form Recognizer Setup guide.

Steps​

Build your project​

1. Add the Azure Form Recognizer Scope to your project.

2. In the Subscription Key and Endpoint properties, enter the Key1 and Endpoint values that you noted when creating your Azure Form Recognizer resource.

3. Add the Train Model activity inside the Azure Form Recognizer Scope activity.

4. Enter the Azure Storage Blob Uri of the blob container that contains the source form documents you're using for your model.

   • For more information, see the Get shared access signature section in the Setup guide.

5. Create and enter a String variable in the Model Id property (e.g., modelId).

   

6. Add the Get Model Info.

7. In the Model Id property, enter the String variable you created for the Train Model activity (e.g., modelId).

8. Create and enter a KeysResult[] variable in the Keys poperty (e.g., keys).

9. Add a For Each activity after the Get Model Info activity.

10. In the TypeArgument property, select String.

11. In the Values property, enter the KeysResult[] variable with an array item identifier and the Keys property (e.g., keys(0).Keys).

12. Inside the For Each activity, add a Write Line activity.

    1. In the Text property, enter "Key: "+item.

13. Add the Delete Model activity after the For Each activity.

14. In the Model Id property, enter the String variable you created for the Train Model activity (e.g., modelId).

15. Run the sequence and verify the Output window includes the expected values.