"I woke up this morning wondering: 'What would happen if I put the Monster Manual of D&D 3.5 in a blender to make a LoRa.'

PonyXL version - use 1.2 strenght, use <lora:ADD_3.5_Monster_Manual_PonyXL:1.2> text, character name, character_profile trigger words.

I first generate txt to img DPM++ 2m, 20 sampling steps.

Then i send the image in img to img and go to 40 sampling steps with ADetailer if need.

Here's what it gives. This LoRa is strange, experimental, but I find it interesting. It's not perfect (that's not the goal), but I think there's a lot to do and test with to achieve great results. Or simply be surprised by the creations of AI?

Test, have fun, and, most importantly, share the strange or not-so-strange results you get!

描述:

训练词语: