While I do love reading Manhua, and I love seeing what you're offering to the users, having tried a few of your lora, I can tell you that they're nearly identical in facial structure. Perhaps train them with different styles to give some alternate variations? Just feedback.