Train your Quick Capture solution

There are two levels of training in Quick Capture; document type training and field extraction training. Both use the same icons to indicate whether the training results are sufficient.

During configuration it is possible that the training circle icon differs between Solution Documents and Field Extraction tabs. This is common during configuration, but it is the goal for a solution to achieve green training icons for all document types on both the Solution Documents and Field Extraction tabs.

The following icons are possible in each situation.

Icon	Solution Documents	Field Extraction
	The training circle icon on the Solution Documents page when a document type has not been trained. This icon is updated automatically to one of the colors below after adding sample documents and then clicking  Train.	The training circle icon before any extraction fields are added to the solution. As soon as you add a field, training occurs automatically and the training circle icon is updated to one of the colors.
	There are insufficient training documents. Each document type should have at least five documents. How to fix: Add additional training documents.	There are not enough confident fields in the training documents. How to fix: Add existing training fields to additional documents. Alternatively, update the Keywords if the label for a field that was not found differs from the original field.
	One or more of the uploaded documents has insufficient quality to be used for document type training. Remove any poor quality documents and replace them with better sample documents. How to fix: Remove poor quality sample documents and add high quality sample documents.	There is at least one training conflict. The affected documents are overlaid with a warning  icon. Hover over the icon to view details about the conflict. How to fix: Review the training conflicts and remove any that are causing problems. This may include removing a sample document if it has a poor quality.
	There are enough sample documents for the document type.	There is sufficient training information for high quality extraction of the configured fields.