After a few months of learning AI as a beginner, I trained some LORA models myself and combined a few models. I feel pretty good about the results and wanted to share them with everyone. Some model combinations worked really well, but since those models don’t allow for sharing after being merged, I can only share this one. After adding my own LORA, the character's skin became fairer and the facial features leaned more towards East Asian appearances. The hands and feet are also less prone to errors. I combined four models for this, and I want to thank the creators of these four models.

3010nc-xx-MixPony

V6.6: Fusion of versions 4, 5, and 6 into the newly trained Lora.

Recommended Sampler: Euler A Steps:20 CFG:7

Positiv Prompt: "score_9, score_8_up, score_7_up, masterpiece, best quality, 8k," Negative Prompt: "score_6, score_5, score_4, lowres, bad quality,"