Description:
This worflow enables upscaling in two different ways:
- Automatically with the Florence module
- Manually with your own prompt
Resources you need:
📂Files for "base" version :
Model : flux1-dev-fp8.safetensors or flux1-dev-fp16.safetensors
in ComfyUI\models\diffusion_models
CLIP : clip_l.safetensors
in ComfyUI\models\clip
📂Files for GGUF version :
Model : Q8, Q6, Q5, Q4, Q3
in ComfyUI\models\unet
CLIP : Q8, Q6, Q5, Q4, Q3
in ComfyUI\models\clip
📂Common Files
Text encoder : t5xxl_fp8_e4m3fn.safetensors or ViT-L-14-TEXT-detail-improved-hiT-GmP-TE-only-HF.safetensors
VAE : ae.safetensors
in ComfyUI\models\vae
ANY upscale model :
Realistic : RealESRGAN_x4plus.pth
Anime : RealESRGAN_x4plus_anime_6B.pth
in ComfyUI\models\upscale_models
📦Custom Nodes :
Description
Optimize for 1080p screen and some correction
FAQ
Comments (1)
This workflow is great! The upscaler is beautiful, it can make eyes and hands much better along with other details. Is there a slider to adjust how much the image is altered during the upscale?
I've noticed when using a character trained lora, after the upscale, their features change a bit too much to retain the source character style. Thanks!