--- language: - en license: creativeml-openrail-m tags: - stable-diffusion - stable-diffusion-diffusers - text-to-image - safetensors inference: true thumbnail: https://huggingface.co/6DammK9/AstolfoMix-SD2/resolve/main/240240-802596570-1536-864-6-192-20240114013047.jpg widget: - text: >- aesthetic, quality, 1girl, boy, astolfo example_title: example 1girl boy #datasets: #- Crosstyan/BPDataset library_name: diffusers --- # AstolfoMix-SD2 (Baseline / Extended / Reinforced / 210b) # - **Based from SD2.1.** I know SD2.1 is hopeless, but it is the best environment for a PoC and to improve my idea. - ["All models are wrong but some are useful."](https://en.wikipedia.org/wiki/All_models_are_wrong) - See [simplified description in CivitAI](https://civitai.com/articles/3409/diagram-of-unet-and-exotic-merging-methods) and [full article in Github](https://github.com/6DammK9/nai-anime-pure-negative-prompt/blob/main/ch05/README_SD2.MD) - *It won't look great, but at least you won't get Miku Hatsune when prompting Suzumiya Haruhi, or having a black hair boy when prompting Link.* ## "210b" ## - Special case of "Add difference". Merge of "Extended" and "Reinforced". - 12 UNETs with 4 CLIPs. - Current version: `210b-AstolfoMix-211209b.safetensors` - Recommended version: "210b" - Recommended CFG: 6.0 ![240240-802596570-1536-864-6-192-20240114013047.jpg](https://huggingface.co/6DammK9/AstolfoMix-SD2/resolve/main/240240-802596570-1536-864-6-192-20240114013047.jpg) ``` parameters (aesthetic:0), (quality:0), (solo:0), (anime:0), (boy:0), (construction helmet:0.98), [[jeans]], [[braid]], [astolfo], [[new york]] Negative prompt: (worst:0), (low:0), (bad:0), (exceptional:0), (masterpiece:0), (comic:0), (extra:0), (lowres:0), (photorealisitc:0) Steps: 192, Sampler: Euler, CFG scale: 6, Seed: 802596570, Size: 1024x576, Model hash: 13696ee702, Model: 210b-AstolfoMix-211209b, VAE hash: df3c506e51, VAE: vae-ft-mse-840000-ema-pruned.ckpt, Denoising strength: 0.7, Clip skip: 2, FreeU Stages: "[{\"backbone_factor\": 1.1, \"skip_factor\": 0.9}, {\"backbone_factor\": 1.2, \"skip_factor\": 0.2}]", FreeU Schedule: "0.0, 1.0, 0.0", FreeU Version: 2, Hires upscale: 1.5, Hires upscaler: Latent, Dynamic thresholding enabled: True, Mimic scale: 1, Separate Feature Channels: False, Scaling Startpoint: MEAN, Variability Measure: AD, Interpolate Phi: 0.3, Threshold percentile: 100, Version: v1.7.0 ``` ## Reinforced ## - Using AutoMBW (bayesian merger but less powerful) for 10 models. - Current version: `209b-AstolfoMix-207b_208b.safetensors` - Recommended version: "209b" - Recommended CFG: 6.0 ![240126-372021954-1536-864-6-256-20240111000506.jpg](https://huggingface.co/6DammK9/AstolfoMix-SD2/resolve/main/240126-372021954-1536-864-6-256-20240111000506.jpg) ``` parameters (aesthetic:0), (quality:0), (anime:0), (race queen:0.98), [[braid]], [[bulge]], [astolfo], [[[[nascar, nurburgring]]]] Negative prompt: (worst:0), (low:0), (bad:0), (exceptional:0), (masterpiece:0), (comic:0), (extra:0), (lowres:0), (photorealisitc:0), (breasts:0.5) Steps: 256, Sampler: Euler, CFG scale: 6, Seed: 372021954, Size: 1024x576, Model hash: 510ede6f03, Model: 209b-AstolfoMix-207b_208b, VAE hash: df3c506e51, VAE: vae-ft-mse-840000-ema-pruned.ckpt, Denoising strength: 0.7, Clip skip: 2, FreeU Stages: "[{\"backbone_factor\": 1.1, \"skip_factor\": 0.9}, {\"backbone_factor\": 1.2, \"skip_factor\": 0.2}]", FreeU Schedule: "0.0, 1.0, 0.0", FreeU Version: 2, Hires upscale: 1.5, Hires steps: 64, Hires upscaler: Latent, Dynamic thresholding enabled: True, Mimic scale: 1, Separate Feature Channels: False, Scaling Startpoint: MEAN, Variability Measure: AD, Interpolate Phi: 0.3, Threshold percentile: 100, Version: v1.7.0 ``` ## Baseline ## - **Uniform merge** of 10 UNETs and 4 CLIPs. - Current version: `209-AstolfoMix-207208-203te.safetensors` - Recommended version: "209-203te" - Recommended CFG: 6.0 - *Prompt is slightly changed from SD1 (anime / photorealistic).* - Intermediate models will be omitted. *This time I'm not merged in sequence, therefore it is not meaningful to show the unuseable models.* ![231236-802596511-1536-864-6-256-20231231171716.png](https://huggingface.co/6DammK9/AstolfoMix-SD2/resolve/main/231236-802596511-1536-864-6-256-20231231171716.png) ``` parameters (aesthetic:0), (quality:0), (anime:0), (solo:0), (boy:0), [niqab], [[hijab]], [astolfo], [[afghanistan]] Negative prompt: (worst:0), (low:0), (bad:0), (exceptional:0), (masterpiece:0), (comic:0), (extra:0), (lowres:0), (photorealisitc:0) Steps: 256, Sampler: Euler, CFG scale: 6, Seed: 802596511, Size: 1024x576, Model hash: a85153fd84, Model: 209-AstolfoMix-207208-203te, VAE hash: df3c506e51, VAE: vae-ft-mse-840000-ema-pruned.ckpt, Denoising strength: 0.7, Clip skip: 2, FreeU Stages: "[{\"backbone_factor\": 1.1, \"skip_factor\": 0.9}, {\"backbone_factor\": 1.2, \"skip_factor\": 0.2}]", FreeU Schedule: "0.0, 1.0, 0.0", FreeU Version: 2, Hires upscale: 1.5, Hires steps: 64, Hires upscaler: Latent, Dynamic thresholding enabled: True, Mimic scale: 1, Separate Feature Channels: False, Scaling Startpoint: MEAN, Variability Measure: AD, Interpolate Phi: 0.3, Threshold percentile: 100, Version: v1.7.0 ``` ### Major difference of methodology ### - **Model selection.** *NP Hard*. Still 12 UNETs, but filtered from 24 models found. Most models are unuseable with its UNET / CLIP / VAE combo. - This time the CLIP is **a merge of 4 CLIPS** instead of the original SD CLIP. I thought the SD2.1's CLIP is bad in performance, but in fact it is still decent, but I found some more useful and merge them together. - Builtin VAE will be `kl-f8-anime2.ckpt`. But I'm still using `vae-ft-mse-840000-ema-pruned.ckpt` for WebUI. Choose what you want. - Minimal idea will be "Replicant-V3 UNET + WD1.5B3 CLIP". The merge ratio made me find more SD2 models in a sophisticated way. ## Receipe ## |Index|Model name (with URL)|UNET used?|CLIP used?| |---|---|---|---| |`_201`|[AllWorkForkRowk](https://civitai.com/models/136088?modelVersionId=233682)||| |`_202`|[Artius V2.1 NSFW](https://civitai.com/models/15975/artius-v21-nsfw?modelVersionId=33119)|Y|| |`_203`|[E621 Rising v2](https://civitai.com/models/32423/e621-rising-v2-a-stable-diffusion-21-furry-model?modelVersionId=38862)||| |`_204`|[hakoMayD](https://civitai.com/models/72668/hakomayd?modelVersionId=77394)|Y|| |`_205`|[Illuminati Diffusion v1.1](https://civitai.com/models/11193/illuminati-diffusion-v11?modelVersionId=13259)||| |`_206`|[Mishi Anime](https://civitai.com/models/7742/mishi-anime?modelVersionId=9125)|Y(a)|| |`_207`|[NijiDiffusion](https://civitai.com/models/55495/nijidiffusion?modelVersionId=59885)||| |`_208`|[Plat Diffusion v1.3.1](https://huggingface.co/p1atdev/plat-diffusion)||| |`_209`|[PVC v4](https://huggingface.co/p1atdev/pvc-v4/tree/checkpoints)||| |`_210`|[Quattro4Merge+i](https://civitai.com/models/11420/quattro4mergei?modelVersionId=13518)|Y|| |`_211`|[Replicant-V3.0](https://civitai.com/models/10701?modelVersionId=75540)|Y|| |`_212`|[Pony Diffusion](https://civitai.com/models/95367/pony-diffusion-v5)||| |`_213`|[Cool Japan Diffusion 2.1.2](https://huggingface.co/aipicasso/cool-japan-diffusion-2-1-2)||Y| |`_214`|[WD 1.5 Beta 2](https://huggingface.co/waifu-diffusion/wd-1-5-beta2)||| |`_215`|[WD 1.5 Beta 3](https://huggingface.co/waifu-diffusion/wd-1-5-beta3)||Y| |`_216`|[YiffAI](https://civitai.com/models/3684/yiffai-2322?modelVersionId=6253)||| |`_217`|[Stable Diffusion v2-1](https://huggingface.co/stabilityai/stable-diffusion-2-1)||Y| |`_218`|[RuminationDiffusion](https://huggingface.co/JosephusCheung/RuminationDiffusion)|Y|| |`_219`|[Scream-SemiRealistic](https://civitai.com/models/99040/sd2scream-semirealistic?modelVersionId=140498)|Y|| |`_220`|[Realgar-v2.1](https://civitai.com/models/70713?modelVersionId=111486)|Y|Y| |`_221`|[RheaSilvia](https://civitai.com/models/68163?modelVersionId=78114)|Y|| |`_222`|[MuaccaMix](https://civitai.com/models/26087/muaccamix?modelVersionId=112322)|Y|| |`_223`|[hakoMayBoy](https://huggingface.co/852wa/hakoMay/blob/main/hakoMayBoy.safetensors)|Y|| |`_224`|[Hurricane](https://civitai.com/models/204162?modelVersionId=240829)|Y(a)||