REAL_pAIn

I spend most of my time working on my SDXL models. However, I learned a lot by working on the SD 1.5 models first. This model takes months of merges and then adds a layer of my custom trained LoRA's on top. Because of the training method used, you may have to dial down the strength on character LoRA's to avoid a burnt in look. Model performs better for use with prompts than with character specific LoRA's based on my testing.

The model was designed for a realistic photo look with improved details. It does have decent flexibility. Hands and feet can require proper prompting to avoid poor quality at times.

There are many excellent SD1.5 models. This one can give a fresh take at times. Sometimes better and sometimes not. Give it a try if you are inclined and please share any great images you produce.

I may train this one some more in the future, but my main focus remains on my SDXL models which I will link in the resources.

All standard disclaimers apply. Use legally and responsibly.