Hey there welcome to my image repo, Today I decided I wanted to compare different cfg and denoise values between models to see how each model is influenced differently by different values. I used the same settings for each generation aside from cfg, denoise, model, and potentially Vae if a model has one baked in (will be noted on image) I will be updating this repo as time goes on and as I process more models, if there is a model you want me to process please just ask!
Note: because of standardized testing methods not all models are going to be utilized to their full potential. I may add seperate images for optimized use of certain models. These models will have the tag OPT.
Prompt:(highres:1.2), beautiful lighting, masterpiece, best quality, intricate, (high detail:1.2), high sharpness, 8k, (1girl), volumetric lighting, vibrant colors, ultra detailed, full body, Sakuya izayoi, Sakuya Izayoi from Touhou Project, apron, blue dress, blue eyes, blush, bowtie, braid, colored shirt, dress, frills, green bow, green bowtie, grey_colored_hair, hair bow, puffy sleeves, puffy short sleeves, shirt, short hair, short sleeves, standing, twin braids, white apron, white shirt, blue_maid_outfit, maid, maid headdress, white_maid_headdress
Negative: bad_prompt_version2, bad-artist, (low resolution:1.2), low quality, bad anatomy, bad hands, missing face, missing head, cut off head
Sampler: Euler A
Steps: 25
Dimensions 640x640 (upscaled to 1280x1280)
Seed: -1
Upscaler: 4xAnimeSharp
- Downloads last month
- 0