I was not satisfied with other LoRA's that are present. One of them was trained on colab, while other doesn't work well with style loras because of underscore tagging the author used.

This one works well with style loras. Prompting instructions are in the model version description (or you can copy from images)