This is a large model trained with the visual works of Eurasia in various periods, including game CG, animation opening, RPG titles, and even a mixture of animation titles from various countries in ancient times, so you can try various animations and games of different eras The style of the title, you only need the trigger word: A visual title screen, you can freely create the anime and game title posters you want

Because of the search for material, it is not recommended for commercial use

Recommended parameters:

Euler a

Steps: 30-70

512x512 of 512x1024

R-ESRGAN 4x+Anime

Denoising: 0.4

And trigger words and non-trigger have 2 effects

One is to add the character lora when it is triggered, and the character lora1 and her name will directly become the title

The other is to not add trigger words, the characters will appear, but adding other tags related to trigger words can also achieve two simultaneous existence

Actually, you made no mistake, I left the watermark. Because the watermarked material will not affect the overall aspect of the work created by AI drawing, especially for this kind of model and language that mainly describes the title. Keeping the watermark allows me to see all kinds of small details randomly drawn by AI. It doesn’t hurt. I need the watermark to simulate the date or subtitle under the game animation title in the AI’s idea to increase the details and mystery of the work and digital watermark It can make people feel the lapse of time in the painting. Others make it easier to confirm that this is a model that cannot be used commercially. If you really can’t bear the problem of watermark, you can choose to remove the tag of the following text in the negative prompt or use the specified plug-in, PS: because I put a lot of watermark images, so every time I generate them, they will not be all is watermarked