Trained on 152 screenshots of them from the anime. It is now better because of a new dataset, and It was trained on LoCon instead Of Lora
Credit goes to Cybermeow for the program used to gather the dataset
Description
FAQ
Comments (1)
This lora of rams is also a great lora. It reproduces the original animation very well and is responsive to prompts. Compared to Emilia's lora and Ram's lora, rem's lora was a little more difficult. Sometimes there were two heads etc. However, increasing the number of generations solved the problem and I am not requesting anything.
I hope to see these kinds of wonderful loras in konno yuuki, leafa, etc (difficulty in pony's prompt's generation)., but I am not paying for anything, so my only comment is that if you are looking for something to create next, please consider it.