


Workflow Name: sd3.5 Model SUPIR High-Definition Upscaling and Restoration
【Workflow Introduction】
This workflow requires you to upload an image you like at the image loading section. It will then use Runninghub Captioner to reverse-generate prompts, combine them with the sd3.5 model for text-to-image generation, pass them to the sampler, and finally upscale and restore the image to high definition using the SUPIR model, resulting in a high-quality image.
【Usage Scenarios】
Simply upload an image to immediately experience this process. Try out your favorite style options and create unique image effects! Give it a try now!
【Key Nodes】
Runninghub Captioner
【Model Version】
sd3.5
Model Name: sd3.5_large.safetensors
SUPIR
Model Name: SUPIR v0Q.ckpt
【LoRA Model】
None
【ControlNet Application】
None
【K Sampler】
CFG: 5.5
Sampling Method: euler
Scheduler: sgm uniform
Denoising: 1


Workflow Name: sd3.5 Model SUPIR High-Definition Upscaling and Restoration
【Workflow Introduction】
This workflow requires you to upload an image you like at the image loading section. It will then use Runninghub Captioner to reverse-generate prompts, combine them with the sd3.5 model for text-to-image generation, pass them to the sampler, and finally upscale and restore the image to high definition using the SUPIR model, resulting in a high-quality image.
【Usage Scenarios】
Simply upload an image to immediately experience this process. Try out your favorite style options and create unique image effects! Give it a try now!
【Key Nodes】
Runninghub Captioner
【Model Version】
sd3.5
Model Name: sd3.5_large.safetensors
SUPIR
Model Name: SUPIR v0Q.ckpt
【LoRA Model】
None
【ControlNet Application】
None
【K Sampler】
CFG: 5.5
Sampling Method: euler
Scheduler: sgm uniform
Denoising: 1