It is recommended to use this LoRA with the distilled (non-base) model.Use the following in your prompt:
quilm style
You may also need to specify the image is an illustration. The LoRA supports high resolutions.
quilm_v3_epoch29.safetensors