Please like and upload your images, I would love to see your creations.
Dataset improved.
This time I generated all images with automatic1111.
Use this model with:
CFG scale: 7-8
Lora intensity: 0.7
Highly recommended to use highres.fix to 1.5x using one of this options:
upscaler: none or any you like with denoise: 0.5 (closer to original but less details)
upscaler: any latent with denoise: 0.5 or higher (more details but changes slightly)
I generated the images using comfyUI.
Use with LoRa strength 0.7
NOTE: In comfyUI the sampler I use for most of the images is dpmpp_3m_sde_gpu, which is not in the list of civitAI when filling in the details for the parameters. I had to chose something so I choose DPM++ 2M SDE Karras.