EtherMix / README.md
Jogn Titor
Create README.md
b67c482
|
raw
history blame
3.71 kB
☁️ Hello, this is Ether Blu Mix 🔵
This is a simple mix aimed at providing a semi-realistic anime style while retaining the knowledge trained on Danbooru-styled tags. Though its base model, AOM3, is fit for anime styled illustrations it's capable of other aesthetics with the correct style of prompting.
The mix also biases towards females and feminine faces, likely due to their prominence within the models as well as the anime aesthetic leaning towards feminine\androgynous faces.
▼ Generation Guidelines
Use whatever sampler, steps, cfg you prefer.
Sample images were generated with
Sampler: DPM++ 2M Karras
Steps: 28
CFG: 7
Clip Skip: 1
*WARNING*
Please input "NSFW" in the negative prompt to reduce instances of unprompted nudity. Due to the models used within the merge there's a bias in generating NSFW imagery.
Negative Prompt Example:
"NSFW, (worst quality, low quality:1.3), watermark, signature"
Most sample images were further upscaled using AUTO1111 Hi-res fix. Please utilize it if you wish to achieve similar results.
Latent (nearest-exact)
Upscale by: 2
Denoising strength: 0.55
Prompting Advice:
Use short prompts mixing prose and danbooru-styled tags. Experiment and give the AI some freedom. When in doubt, change up your prompting style.
General guidelines:
Prompt in anime\danbooru-style models
More anime style images
Prompt in photography\non-anime models
More realistic images
The sample images were generated using Waifu Diffusion VAE kl-f8-anime2. This VAE is not mandatory. Please use whichever VAE you prefer.
▼ Model Merging
I have no training or real knowledge on machine learning. I merged the model using the information provided on WarriorMama777's AOM model page. I very arbitrarily used the UBM weights given to merge and create AOM3 and its derivatives and modified the values with trial and error until I was visually happy with it.
I utilized the SuperMerger extension for AUTO1111 by hako-mikan.
▼ Merging
MODEL A: AOM3.safetensors
MODEL B: Chilloutmix_fp16.safetensors
Weight Sum merge - Use MBW
Alpha: 1
0,0.3,0.3,0.3,0.3,0.3,0,0,0,0,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0,0
Model: AOM3-ChilloutSumS2-03
MODEL A: AOM3-ChilloutSumS2-03
MODEL B: deliberate_v2.ckpt
Weight Sum merge - Use MBW
Alpha: 1
0,0.2,0.2,0.2,0.2,0.2,0,0,0,0,0.2,0.1,0.2,0.2,0.2,0.2,0.2,0.1,0.1,0.1,0.2,0.2,0.1,0.2,0.2,0.2
Model: AOM3-ChilloutSumS2-03-DelibV2SumS1-02
MODEL A: AOM3-ChilloutSumS2-03-DelibV2SumS1-02
MODEL B: pastelmix-better-vae-fp16.safetensors
Weight Sum merge - Use MBW
Alpha: 1
0,0.2,0.2,0.2,0.2,0.2,0,0,0,0,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0.2,0,0
Final Model: EBM-v1
Models Used:
AOM3.safetensors
D124FC18F0232D7F0A2A70358CDB1288AF9E1EE8596200F50F0936BE59514F6D
chilloutmix_NiPrunedFp16Fix.safetensors
59FFE2243A25C9FE137D590EB3C5C3D3273F1B4C86252DA11BBDC9568773DA0C
deliberate_v2.ckpt
9ABA26ABDFCD46073E0A1D42027A3A3BCC969F562D58A03637BF0A0DED6586C9
pastelmix-better-vae-fp16.safetensors
D01A68AE76F97506363F387F5F28BB564AD9E20924844FD5945E600B72D39E79
Credits:
Thank you to WarriorMama777 for providing AOM and other various mixes as well as detailing your workflow and inspiring me to try mixing my own models.
Thank you to hako-mikan for creating the SuperMerger extension allowing a quicker workflow for merging.
Thank you to the creator of ChilloutMix.
Thank you to XpucT for creating Deliberate.
Thank you to andite for Pastel-Mix.
Thank you to AUTO1111 for creating the web UI everyone uses.
Thank you to StabilityAI for starting everything with Stable Diffusion.
Thank you to the entire SD community for continuing to openly share and create.