How could I get more costumer TTS results based on pretrained models?

Hi I am new onboarding to mozilla TTS. I try to test the result by pretrained model + CMD. How could i get the result sounds younger and energic? Thanks!