I recommend experimenting with LoRA from 0.25 to 1 + You can also try other eph LoRA
Doomer-style.safetensors
1107161_training_data.zip