Step 1: Go to https://app.nanonets.com and select "Build a document type separator"
Step 2: Specify the different document types or "classes"
Step 3: Add the specific document images/pages for that particular class and press
"Train" after the last class
Step 4: A model will be trained and you can view the Metrics associated with model training and performance under the "Model Metrics" screen
Step 5: View results by uploading unseen images in the test screen. In the JSON response, the document type/class with the highest score (generally >0.9) is the predicted class.
