This Lora model has been trained using a dataset of 100 images, providing a solid foundation for generating high-quality images. For the best results, it is recommended to use the ReV Animated and MeinaMix models, which offer superior image resolution.
Recommended Configuration:Clip Skip: Set to 2.
Sampling Steps: Minimum 30.
Recommended Dimensions: Minimum 768x768, maximum 1024x1024 for the first generation of the image. Subsequently, you can scale the image using the extras or img2img features.
SD VAE: It is recommended to use orangemix.vae.pt or vae-ft-mse-840000-ema-pruned | 840000 | 840k SD1.5 VAE to further enhance the quality of the generated images.
Happy creating!