Stable Diffusion 3.5 Large gacha with LoRA Detailer UpScaler ControlNet v01
created 4 months ago
img2img
lora
controlnet
upscale
129 nodes
1.1k
177
Outputs
.webp)




.webp)
Description
Stable Diffusion 3.5 Large gacha with LoRA Detailer UpScaler ControlNet v01
Anime style preset.
Stable Diffusion 3.5 Large using large amount of VRAM.
I recommended over 24 GB VRAM.
ControlNet is HEAVY IMPACT to VRAM
Enable 1ControlNet = +8GB VRAM
Enable 2ControlNets = +16GB VRAM ( I Cannot Execute... )
Enable 3ControlNets = +24GB VRAM ( I Cannot Execute... )
You can Enable / Disable Group nodes from "Fast Groups Bypasser (rgthree)" node.
Download models from
Comfy-Org/stable-diffusion-3.5-fp8
https://huggingface.co/Comfy-Org/stable-diffusion-3.5-fp8/tree/main
stabilityai/stable-diffusion-3.5-controlnets
https://huggingface.co/stabilityai/stable-diffusion-3.5-controlnets
Stable Diffusion Diffusion 3.5 Large ガチャ LoRA Detailer UpScaler ControlNet v01
アニメスタイルプリセットですが SD3.5 はあまり アニメ系が得意でない模様。
Stable Diffusion 3.5 Large なので、大量のVRAMを使用します。
24 GB 以上の VRAM を推奨します。
ControlNet は VRAM に大きな影響を及ぼします
1 ControlNet = +8GB VRAM
2 ControlNet = +16GB VRAM ( 実行できていません)
3 ControlNet = +24GB VRAM (実行できていません)
"Fast Groups Bypasser (rgthree)" ノードから グループ一括で ノードを有効/無効にできます。
モデルは以下からダウンロードできます。
Comfy-Org/stable-diffusion-3.5-fp8
https://huggingface.co/Comfy-Org/stable-diffusion-3.5-fp8/tree/main
stabilityai/stable-diffusion-3.5-controlnets
https://huggingface.co/stabilityai/stable-diffusion-3.5-controlnets
Built-in nodes
EmptySD3LatentImage
Custom nodes
LoraLoader
MathExpression|pysssss
UltralyticsDetectorProvider
BboxDetectorSEGS
DetailerForEachPipe
VAEDecodeTiled
Seed Everywhere
ToBasicPipe
String Literal
ImpactFloat
PreviewImage
CheckpointLoaderSimple
LoadImage
ControlNetLoader
ControlNetApplyAdvanced
Int Literal
ImpactInt
JoinStringMulti
Scheduler Selector
ImpactSchedulerAdapter
ImpactKSamplerAdvancedBasicPipe
CLIPLoader
CR Upscale Image
FromBasicPipe
UltimateSDUpscale
UpscaleModelLoader
ImpactImageBatchToImageList
SaveImage
Textbox
SizeFromPresetsSDXL
Sampler Selector
WD14Tagger|pysssss
Canny
LeReS-DepthMapPreprocessor
CLIPTextEncode