Thank you for your interesting project, I'm wondering how can I get personality scores of plain text with the trained model?