SegMoE-2x1-v0: Segmind Mixture of Diffusion Experts

SegMoE-2x1-v0 is an untrained Segmind Mixture of Diffusion Experts Model generated using segmoe from 2 Expert SDXL models. SegMoE is a powerful framework for dynamically combining Stable Diffusion Models into a Mixture of Experts within minutes without training. The framework allows for creation of larger models on the fly which offer larger knowledge, better adherence and better image quality.

Usage

This model can be used via the segmoe library.

Make sure to install segmoe by running

pip install segmoe

from segmoe import SegMoEPipeline

pipeline = SegMoEPipeline("segmind/SegMoE-2x1-v0", device = "cuda")

prompt = "cosmic canvas, orange city background, painting of a chubby cat"
negative_prompt = "nsfw, bad quality, worse quality"
img = pipeline(
    prompt=prompt,
    negative_prompt=negative_prompt,
    height=1024,
    width=1024,
    num_inference_steps=25,
    guidance_scale=7.5,
).images[0]
img.save("image.png")

Config

Config Used to create this Model is:

base_model: SG161222/RealVisXL_V3.0
num_experts: 2
moe_layers: all
num_experts_per_tok: 1
experts:
  - source_model: frankjoshua/juggernautXL_v8Rundiffusion
    positive_prompt: "aesthetic, cinematic, hands, portrait, photo, illustration, 8K, hyperdetailed, origami, man, woman, supercar"
    negative_prompt: "(worst quality, low quality, normal quality, lowres, low details, oversaturated, undersaturated, overexposed, underexposed, grayscale, bw, bad photo, bad photography, bad art:1.4), (watermark, signature, text font, username, error, logo, words, letters, digits, autograph, trademark, name:1.2), (blur, blurry, grainy), morbid, ugly, asymmetrical, mutated malformed, mutilated, poorly lit, bad shadow, draft, cropped, out of frame, cut off, censored, jpeg artifacts, out of focus, glitch, duplicate, (airbrushed, cartoon, anime, semi-realistic, cgi, render, blender, digital art, manga, amateur:1.3), (3D ,3D Game, 3D Game Scene, 3D Character:1.1), (bad hands, bad anatomy, bad body, bad face, bad teeth, bad arms, bad legs, deformities:1.3)"
  - source_model: SG161222/RealVisXL_V3.0
    positive_prompt: "cinematic, portrait, photograph, instagram, fashion, movie, macro shot, 8K, RAW, hyperrealistic, ultra realistic,"
    negative_prompt: "(octane render, render, drawing, anime, bad photo, bad photography:1.3), (worst quality, low quality, blurry:1.2), (bad teeth, deformed teeth, deformed lips), (bad anatomy, bad proportions:1.1), (deformed iris, deformed pupils), (deformed eyes, bad eyes), (deformed face, ugly face, bad face), (deformed hands, bad hands, fused fingers), morbid, mutilated, mutation, disfigured"

Other Variants

We release 3 merges on Hugging Face,

• SegMoE 4x2 has four expert models.
• SegMoE SD 4x2 has four Stable Diffusion 1.5 expert models.

Comparison

The Prompt Understanding seems to improve as shown in the images below. From Left to Right SegMoE-2x1-v0, SegMoE-4x2-v0, Base Model (RealVisXL_V3.0)

three green glass bottles
panda bear with aviator glasses on its head
the statue of Liberty next to the Washington Monument

Model Description

• Developed by: Segmind
• Developers: Yatharth Gupta and Vishnu Jaddipal.
• Model type: Diffusion-based text-to-image generative mixture of experts model
• License: Apache 2.0

Out-of-Scope Use

The SegMoE-2x1-v0 Model is not suitable for creating factual or accurate representations of people, events, or real-world information. It is not intended for tasks requiring high precision and accuracy.

Advantages

• Benefits from The Knowledge of Several Finetuned Experts
• Training Free
• Better Adaptability to Data
• Model Can be upgraded by using a better finetuned model as one of the experts.

Limitations

• Though the Model improves upon the fidelity of images as well as adherence, it does not be drastically better than any one expert without training and relies on the knowledge of the experts.
• This is not yet optimized for speed.
• The framework is not yet optimized for memory usage.

Citation

@misc{segmoe,
  author = {Yatharth Gupta, Vishnu V Jaddipal, Harish Prabhala},
  title = {SegMoE},
  year = {2024},
  publisher = {HuggingFace},
  journal = {HuggingFace Models},
  howpublished = {\url{https://huggingface.co/segmind/SegMoE-2x1-v0}}
}