Train model

Using the source form documents stored in your specified blob storage container (Azure Storage Blob Uri), the Train Model activity calls the Form Recognizer Train Custom Model API to create and train a new custom model.

After completing the train model operation, the activity outputs the identifier assigned to the model as a String (Model Id) that you can use as an input variable in subsequent activities (e.g., Get Model Info).

How it works

The following steps and message sequence diagram is an example of how the activity works from design time (that is, the activity dependencies and input/output properties) to run time.

1. Complete the Setup steps.
2. Add the Azure Form Recognizer Scope activity to your project.
3. Add the Train Model activity inside the Azure Form Recognizer Scope activity.
4. Enter values for the Input properties.
5. Create and enter a String variables for the Output properties.
6. Run the activity.

   • Your input properties are sent to the Train Model API.

   • The API returns the String value to your output property variables.

     

Properties

The values for the following properties are specified when adding this activity to your project in UiPath Studio.

Common

DisplayName

The display name of the activity.

Attributes	Details
Type	String
Required	Yes
Default value	Train Model
Allowed values	Enter a String or String variable.
Notes	N/A

Input

Azure Storage Blob Uri

The shared access signature (SAS) URI of the Azure blob storage container that contains the source form documents to use in the training model.

Attributes	Details
Type	String
Required	Yes
Default value	Empty
Allowed values	Enter a String or String variable.
Notes	To get the URI value, follow the Get shared access signature steps in the Setup guide (for quick reference, see image below). Supported form document formats: PNG, JPEG, and PDF (all other format types are ignored).

Options

Include Subfolders

Indicates the source form documents are stored in a subfolder in the Azure blob storage container.

Attributes	Details
Type	Checkbox
Required	No
Default value	Not Selected
Allowed values	Selected or Not Selected
Notes	By default, the train model operation only uses form documents located in the root of your Azure blob storage container. If your form documents are stored in a subfolder, select this option.

Source Filter Prefix

Limits the training data set to form documents that have a name beginning with the specified value.

Attributes	Details
Type	String
Required	No
Default value	Empty
Allowed values	Enter a String or String variable.
Notes	If your Azure blob storage container contains various form documents, you can specify which ones to include in this training model by entering the shared portion of the document names. For example, if you enter "Invoice", only the form documents with a name beginning with "Invoice" are used.

Use Label File

If selected, the model uses label files.

Attributes	Details
Type	Checkbox
Required	No
Default value	Not Selected
Allowed values	Selected or Not Selected
Notes	N/A

Misc

Private

If selected, the values of variables and arguments are no longer logged at Verbose level.

Attributes	Details
Type	Checkbox
Required	No
Default value	Not Selected
Allowed values	Selected or Not Selected
Notes	N/A

Output

Model Id

The identifier assigned to the training model.

Attributes	Details
Type	String
Required	No (required to use the output data in subsequent activities)
Default value	Empty
Allowed values	Enter a String variable
Notes	The variable you create for this property can be used as an input variable in other Azure Form Recognizer activities (e.g., Get Model Keys, Get Model Info).