license: creativeml-openrail-m
language:
- en
tags:
- General Purpose
- Versatile
- Anime
- Art
- Realistic
- stable-diffusion
- stable-diffusion-diffusers
- diffusers
- text-to-image
- artificialguybr
- pmango300574
pipeline_tag: text-to-image
base_model:
- artificialguybr/freedom
- Yntec/MangledMerge3_768
base_model_relation: merge
Another Fine Mess
This model guarantes better generations than artificialguybr's Freedom.Redmond model when you use Anime in the prompt and better ones than pmango300574's MangledMerge3 when you don't! Those are my favorite models from the Stable Diffusion 2.1 era and it's the first time I make a merge with this architecture, but don't expect anything better than models using Stable Diffusion 1.5, 2.1 did not become popular for a reason. But still, these will be very unique outputs unlike any other image model. Samples and prompts (all use seed 9119):
portrait, (solo, 1boy, 1girl),((JOE BIDEN WITH HIS DAUGHTER)) day time,best quality, (anime, waifu, new, newest:1.2), (exceptional, best aesthetic, new, newest, best quality, masterpiece, extremely detailed, anime, waifu:1.2), anime style, anime, cinematic, film still
A beautiful young girl with blond hair and beautiful eyes, closeup, cute face, professional bike rider fat girl, jersey, glad, beautiful landscape, beautiful, realistic, 4k, perfect composition, masterpiece
Super cute robot, cute, raining, forest, high definition, vignette,
best quality, detailed background, girl, ,euro_street, random_wear.
The Diffusers versions has the FreedomVAE baked in, on A1111 baking it makes it have an identical output, so I don't know what's that about.
Original pages:
https://civitai.com/models/5395?modelVersionId=93316 (Mangled Merge 3)
https://huggingface.co/artificialguybr/freedom
Recipe:
- SuperMerger Weight Sum Use MBW 0,0,0,0,1,1,1,0,0,0,1,1,1,1,0,0,0,1,1,1,0,0,0,1,1,1
Model A:
freedom
Model B:
MangledMerge_v3
Output:
AnotherFineMess