该ComfyUI工作流程支持您挑选非SD模型来创建基础图像,并将其送入配备controlnet的SDXL模型进行增强。各个不同的起始模型带来不同的优势,例如PixArt能够提供丰富的多样性,Aura能够确保快速的精确度,而FLUX则能够提升整体图像质量。您可以通过调整控制网的强度来获得更高的图像保真度或激发更多的创造力,但这可能会改变原始模型的风格。
工作流界面图
作者:mithrillion
模型详情
Checkpoints
900mPixartSigma_base.safetensors
HunyuanDiT-v1.2_ema.pt
aura_flow_0.2.safetensors
zavychromaxl_v90.safetensors👉
节点详情
内置节点
Any Switch (rgthree)
AuraSR.AuraSRUpscaler
DownloadAndLoadFlorence2Model
Fast Groups Muter (rgthree)
Florence2Run
FluxGuidance
Image Comparer (rgthree)
自定义节点
Comfyroll Studio
– CR Apply Multi-ControlNet
– CR SDXL Aspect Ratio
– CR Upscale Image
– CR Multi-ControlNet Stack
ComfyUI
– CLIPTextEncode
– PreviewImage
– VAEDecode
– VAELoader
– DualCLIPLoader
– BasicScheduler
– BasicGuider
– UNETLoader
– CheckpointLoaderSimple
– SamplerCustomAdvanced
– VAEEncode
– SaveImage
– ImageScaleToTotalPixels
– KSampler
– KSamplerSelect
– RandomNoise
ComfyUI Essentials
– SDXLEmptyLatentSizePicker+
Extra Models for ComfyUI
– HYDiTTextEncoderLoader
– HYDiTCheckpointLoader
– T5v11Loader
– PixArtCheckpointLoader
– T5TextEncode
– HYDiTTextEncodeSimple
KJNodes for ComfyUI
– ColorMatch
pythongosssss/ComfyUI-Custom-Scripts
– ShowText|pysssss
Tiled Diffusion & VAE for ComfyUI
– TiledDiffusion
WAS Node Suite
– Text to Conditioning
评论0