I ran the same prompt using many samplers at different steps counts to evaluate which one(s) give a decent quality at a low step count. I have not used the “restore faces” option. Here are my observations related to image quality (artifacts) and convergence. Quality at lower steps At 10 steps, a few samplers are unusable: DPM++ 2M and its variants, DDIM. At 15 steps, all samplers are OK except DPM++ 2M SDE and its Karras variant are unusable....
When you generate a file using ComfyUI, metadata are added to the image automatically. Amongst the metadata, there is the full workflow including the prompt. If you want to remove those data, you can use ImageMagick convert with the --strip option. convert image.png --strip image_strip.png If you want to alter the original file, use mogrify: mogrify --strip image.png
While AUTOMATIC1111 can generate images based on prompt variations, I haven’t found the same possibility in ComfyUI. However, you can achieve the same result thanks to ComfyUI API and curl. When you click “queue prompt” in ComfyUI, it actually sends a POST request with the whole workflow as JSON data to http://127.0.0.1:8188/prompt . To get the workflow as JSON, go to the UI and click on the settings icon, then enable Dev mode Options and click close....
I mostly use the FaceDetailer node to fix eyes and mouths, it really makes a difference, especially with models on CivitAI. You will find the workflow below. Note that on Nvidia GPU, I can use the SAMLoader node with “GPU” but on AMD GPU (RX 6700 XT), I must use “CPU”.
ComfyUI is a web UI to run Stable Diffusion and similar models. It is an alternative to Automatic1111 and SDNext. One interesting thing about ComfyUI is that it shows exactly what is happening. The disadvantage is it looks much more complicated than its alternatives. In this post, I will describe the base installation and all the optional assets I use. I will provide workflows for models you find on CivitAI and also for SDXL 0....