This LORA is useful to create high quality realistic anime images. Trained by using generated images from RaemuXL model as datasets, then ported into Pony Diffusion XL. Only works with Pony Diffusion XL. SDXL models and also SDXL+Pony XL merges don't work. If you're insisting to use it on non-pony SDXL models, please download the RaemuXL checkpoint model instead.

How to use this LORA:

1. Use the keyword "raemu style" to activate it.

2. Strength values ranging from 0.6 to 1.0.

3. Intended to be applied with artwork with original characters within the model.

4. Requires to do prompt matrix first when applied to characters with LORA extensions. Some of them may not applied its effect.

5. If random garbage image appears, try to restart the A1111's SD WebUI first.

Please note that this is very first version, expect some hiccups.