About unseen speakers

Hi. I notice that in the demo page (Section 4.3), you only did parallel voice cloning with unseen speakers, have you tried testing with different text with these unseen speakers?
Besides, is it possible to get some pretrained models? 
Thank you very much. 