Update README.md

c5166e5 verified 5 months ago

21.5 kB

language:
  - en
pipeline_tag: text-to-image
library_name: diffusers
tags:
  - text-to-image
  - stable-diffusion
  - safetensors
  - stable-diffusion-xl
  - merge
license: openrail++

RealCore Illustrious 1.0

RealCore Illustrious is a merged Illustrious checkpoint built from multiple models and LoRAs, designed to deliver stunning results in realistic and photo-realistic styles. Fully supports NSFW generation and integrates seamlessly with LCM acceleration.

📷Realistic & photo-realistic rendering
🔀Complex multi-model merge with LoRAs
⚡Excellent LCM compatibility
🔓NSFW supported

RealCore Illustrious 1.0

SDXL Merge | Base: Illustrious

A merged Illustrious checkpoint built from multiple models and LoRAs, designed to deliver stunning results in realistic and photo-realistic styles. Fully supports NSFW generation and integrates seamlessly with LCM acceleration.

📷 Realistic & photo-realistic rendering
🔀 Complex multi-model merge with LoRAs
⚡ Excellent LCM compatibility
🔓 NSFW supported

RealCore Pony 1.0

SDXL Merge | Base: Pony_Diffusion_V6_XL

A merged Pony Diffusion V6 checkpoint built from multiple models and LoRAs, designed to deliver stunning results in realistic styles. Fully supports NSFW generation and integrates seamlessly with LCM acceleration.

📷 Realistic & photo rendering
🔀 Complex multi-model merge with LoRAs
⚡ Excellent LCM compatibility
🔓 NSFW supported

RealCore XL 1.0

SDXL Merge | Base: SDXL 1.0

A merged Stable Diffusion XL checkpoint built from multiple models and LoRAs, designed to deliver stunning results in realistic and amateur photo styles. Fully supports NSFW generation and integrates seamlessly with LCM acceleration.

📷 Realistic & amateur photo rendering
🔀 Complex multi-model merge with LoRAs
⚡ Excellent LCM compatibility
🔓 NSFW supported

RealCore Illustrious HuggingFace RealCore Pony HuggingFace RealCore XL HuggingFace RealCore Illustrious CivitAI RealCore Pony CivitAI RealCore XL CivitAI

All models are optimized for ComfyUI + LCM and support NSFW generation.

🧩 Model Overview

RealCore Illustrious is a diffusion-based Illustrious checkpoint obtained through a complex merging process of multiple realistic models combined with several custom LoRAs. It produces highly detailed, lifelike, and emotionally grounded results — especially suited for realistic, photo-realistic, nsfw, and professional photo compositions.

Type: SDXL Merge
Base: Illustrious
Supports: Realistic / Photo-realistic / NSFW
Optimized for: ComfyUI + LCM

⚙️ ComfyUI Workflows

Workflow	Description	Link
LCM DMD2	Optimized for fast inference with LCM acceleration	Download LCM DMD2 Workflow

🧠 Prompting

For example prompts, styles, and tag guidance — please check 👉 RealCore Illustrious on CivitAI

🧾 Checkpoint Recipe

RealCore Illustrious Beta Recipe

version 0.1.0
model "illustrious\\cyberrealistic_v70.safetensors" model_config="sdxl-sgm" merge_space="weight"
model "illustrious\\goddessOfRealism_gorV6ilxlVAE.safetensors" model_config="sdxl-sgm" merge_space="weight"
model "illustrious\\damnIllustriousPony_v50IllustriousNoobai.safetensors" model_config="sdxl-sgm" merge_space="weight"
model "illustrious\\truespectraStudio_v10.safetensors" model_config="sdxl-sgm" merge_space="weight"
merge "🔨_weight_sum_with_json" &0 &1 &2 &3 json_params="{
  \"global\": \"1.000,0.000,0.000,0.000\",
  \"CLIP_L\": \"0.250,0.250,0.250,0.250\",
  \"CLIP_G\": \"0.250,0.250,0.250,0.250\",
  \"time_embed\": \"0.250,0.250,0.250,0.250\",
  \"label_emb\": \"0.250,0.250,0.250,0.250\"
}"
merge "geometric_median" &0 &1 &2 &3 eps=1e-06 maxiter=100 ftol=1e-20
literal 1.0 model_config="sdxl-sgm" merge_space="param"
merge "pick_component" &6 "clip_l"
merge "pick_component" &6 "clip_g"
merge "fallback" &7 &8
literal 0.0 model_config="sdxl-sgm" merge_space="param"
merge "pick_component" &10 "vae"
merge "fallback" &9 &11
merge "pick_component" &10 "diffuser"
merge "fallback" &12 &13
merge "weighted_sum" &0 &5 &14
literal 0.5 model_config="sdxl-sgm" merge_space="param"
merge "pick_component" &16 "clip_l"
merge "pick_component" &16 "clip_g"
merge "fallback" &17 &18
merge "fallback" &19 &11
merge "fallback" &20 &13
merge "crossover" &4 &15 &21 0.5
model "illustrious\\fantasticReality_10.safetensors" model_config="sdxl-sgm" merge_space="weight"
model "illustrious\\illustriousXL_v01.safetensors" model_config="sdxl-sgm" merge_space="weight"
model "illustrious\\realcosplay1024px_realcosplayV10Illust.safetensors" model_config="sdxl-sgm" merge_space="weight"
merge "subtract" &25 &24
model "illustrious\\realcosplay1536px_v21.safetensors" model_config="sdxl-sgm" merge_space="weight"
merge "subtract" &27 &24
model "illustrious\\illustriousCSG_unrealengineILProV1.safetensors" model_config="sdxl-sgm" merge_space="weight"
merge "subtract" &29 &24
model "illustrious\\illustriousFacial_v20.safetensors" model_config="sdxl-sgm" merge_space="weight"
merge "subtract" &31 &24
merge "geometric_median" &26 &28 &30 &32 eps=1e-06 maxiter=100 ftol=1e-20
merge "add_difference" &24 &33 1.0
merge "fallback" &34 &24
model "illustrious\\realistic\\Realim_Lora_BSY_IL_V1_RA42.safetensors" model_config="sdxl-kohya_kohya_lora" merge_space="weight"
merge "convert_'sdxl-kohya_kohya_lora'_to_base" &36
merge "convert_'sdxl-kohya'_to_'sdxl-sgm'" &37
model "illustrious\\realistic\\Realistic_furry.safetensors" model_config="sdxl-kohya_kohya_lora" merge_space="weight"
merge "convert_'sdxl-kohya_kohya_lora'_to_base" &39
merge "convert_'sdxl-kohya'_to_'sdxl-sgm'" &40
model "illustrious\\realistic\\ScreencUp.safetensors" model_config="sdxl-kohya_kohya_lora" merge_space="weight"
merge "convert_'sdxl-kohya_kohya_lora'_to_base" &42
merge "convert_'sdxl-kohya'_to_'sdxl-sgm'" &43
merge "🔨_ties_lora_with_json" &38 &41 &44 json_params="{
  \"lambda\": 1,
  \"weights\": \"0.500,0.500,0.500\",
  \"density\": 0.4,
  \"vote_sgn\": true,
  \"apply_stock\": false,
  \"apply_median\": true
}"
merge "add_difference" &35 &45 1.0
merge "fallback" &46 &35
merge "crossover" &23 &47 0.66 1.0
merge "🔨_karcher_mean_with_json" &0 &1 &3 json_params="{
  \"global\": \"0.382,0.382,0.236\",
  \"max_iter\": 30,
  \"tol\": 3e-7
}"
model "illustrious\\realistic\\ALTGRL(CYBERILL).safetensors" model_config="sdxl-kohya_kohya_lora" merge_space="weight"
merge "convert_'sdxl-kohya_kohya_lora'_to_base" &50
merge "convert_'sdxl-kohya'_to_'sdxl-sgm'" &51
model "illustrious\\realistic\\RSGRL-REALLUXURA(10_10_25).safetensors" model_config="sdxl-kohya_kohya_lora" merge_space="weight"
merge "convert_'sdxl-kohya_kohya_lora'_to_base" &53
merge "convert_'sdxl-kohya'_to_'sdxl-sgm'" &54
model "illustrious\\realistic\\KRGRLv2-000003.safetensors" model_config="sdxl-kohya_kohya_lora" merge_space="weight"
merge "convert_'sdxl-kohya_kohya_lora'_to_base" &56
merge "convert_'sdxl-kohya'_to_'sdxl-sgm'" &57
merge "🔨_ties_lora_with_json" &52 &55 &58 json_params="{
  \"lambda\": 1,
  \"weights\": \"0.500,0.500,0.200\",
  \"density\": 0.5,
  \"vote_sgn\": true,
  \"apply_stock\": false,
  \"apply_median\": true
}"
merge "add_difference" &49 &59 1.0
merge "fallback" &60 &49
merge "🔨_karcher_mean_with_json" &22 &48 &61 &2 json_params="{
  \"global\": \"0.000,0.260,0.460,0.280\",
  \"max_iter\": 30,
  \"tol\": 3e-7,
  \"OUT04\": \"0.000,0.157,0.375,0.468\",
  \"OUT05\": \"0.000,0.157,0.375,0.468\",
  \"OUT06\": \"0.000,0.157,0.375,0.468\",
  \"IN\": \"0.000,0.230,0.477,0.294\",
  \"OUT\": \"0.000,0.273,0.312,0.415\",
  \"CLIP_L\": \"0.364,0.091,0.273,0.273\",
  \"CLIP_G\": \"0.364,0.091,0.273,0.273\"
}"
model "illustrious\\realistic\\FLMGR-IL.safetensors" model_config="sdxl-kohya_kohya_lora" merge_space="weight"
merge "convert_'sdxl-kohya_kohya_lora'_to_base" &63
merge "convert_'sdxl-kohya'_to_'sdxl-sgm'" &64
dict BASE=0.0 IN00=0.0 IN01=0.0 IN02=0.0 IN03=0.0 IN04=0.0 IN05=0.0 IN06=0.0 IN07=0.0 IN08=0.0 M00=0.0 OUT00=0.0 OUT01=0.0 OUT02=0.15 OUT03=0.3 OUT04=0.4 OUT05=0.5 OUT06=0.3 OUT07=0.15 OUT08=0.0 VAE=0.0
literal &66 model_config="sdxl-supermerger_blocks" merge_space="param"
merge "convert_sdxl_blocks_to_sgm" &67
merge "add_difference" &62 &65 &68
merge "fallback" &69 &62

RealCore Illustrious v10 Recipe

version 0.1.0
model "ritya\\RealCore_Illustrious_beta.safetensors" model_config="sdxl-sgm" merge_space="weight"
model "illustrious\\cyberrealistic_v70.safetensors" model_config="sdxl-sgm" merge_space="weight"
model "illustrious\\damnIllustriousPony_v50IllustriousNoobai.safetensors" model_config="sdxl-sgm" merge_space="weight"
merge "🔨_karcher_mean_with_json" &0 &1 &2 json_params="{
  \"global\": \"1.000,0.000,0.000\",
  \"max_iter\": 50,
  \"tol\": 1e-7,
  \"IN00\": \"0.143,0.429,0.429\",
  \"IN01\": \"0.189,0.406,0.406\",
  \"IN02\": \"0.211,0.395,0.395\",
  \"IN03\": \"0.250,0.375,0.375\",
  \"IN04\": \"0.311,0.345,0.345\",
  \"IN06\": \"0.360,0.320,0.320\",
  \"IN07\": \"0.390,0.305,0.305\",
  \"IN08\": \"0.399,0.301,0.301\",
  \"IN\": \"0.333,0.333,0.333\"
}"