- Jan 24, 2020
- 1,401
- 3,791
I switched to Forge recently and I don't regret it. Try it out. You can keep using the checkpoint models in the old folder by editing the user.bat file with filepaths for the different folders. Olivio Sarika made an excellent tutorial. If you don't like it then you will simply delete Forge, no harm, no foul.. However if you choose to switch it's easy enough to simply cut/paste everything over to the new folders. My experience is that Forge is much more stable and faster than A1111. It comes with a bunch of very useful extensions that likely is unfamiliar. I would likely not have found these extensions on my own. Forge allowed me to get back into SDXL, that I had given up on because it were so slow and problematic with my old GPU.
- Sep 26, 2018
- 3,040
- 3,118
- Dec 31, 2016
- 316
- 708
This image uses the same workflow, except it's about 1/3 of the nodes as i were testing the advanced clip nodes to test the weight interpretation and have more "a1111 like" prompts to share. So almost all nodes were tripled run the comfy, comfy++ and a1111 interpretations at the same time.
There's still some unused/muted nodes and a lot of nice noodles. Might want to check the collapsed clip nodes as it's currently set to a1111 if there's some odd/unexpected behavior with prompts
Need to tweak the denoising for this image some more but that's irrelevant for the workflow itself.
Edit: forgot to include that the prompt is "stolen" from
You must be registered to see the links
Last edited:
- Dec 31, 2016
- 316
- 708
- Jan 24, 2020
- 1,401
- 3,791
It depends on the GPU you have. If you have a lower end card you will make bigger gains. If you have a high end card you will get, I don't remember exactly but lets say 5-10% perhaps, probably a little less. It's very noticeable for me at least with an old GTX1070. The biggest thing is that Forge is more stable, less bugs and better memory management. Each user case scenario will be different because each PC hardware component makes a difference in the end even though the GPU is the big thing.
- Mar 17, 2021
- 282
- 962
Yes, i say something about Forge here, i'm totally in accord with you. Better experience than a1111.
Actually i don't have time to test a1111 1.8, maybe i try it later, but Forge seem optimized a lot than a1111.
Did you run with same option i wrote in quoted 3ad?
link inside quoted 3ad.
- Jan 24, 2020
- 1,401
- 3,791
You shared settings? No I always find my own. 3ad? I'm not sure what that refers to.
Respectfully I am new to this, just recently installed everything last night - are you able to share what Model and LoRA if any for these renderings? And maybe even the prompt? Really like the appearance. Any help would be appreciated.
Interestingly, I can get decent renders of something like an E60 Chassis BMW 5 series, but something less ubiquitous (like my once loved SAAB 9000 Aero) becomes a jumbled mess. Ditto for aircraft, I am an aviation photographer, and I have tried 3 or 4 different AI image generation tools, and even a simple F-15C comes out as a jumbled mess, with perhaps even a third intake. They also look really Macross which is cool, but I was looking for an F-15, not a VF-1. You guys do some beautiful work here, I will post up my newbie stuff when I go downstairs to my Windows PC. I met Jimwalrus from another Board, and he has been so helpful.
- Dec 21, 2020
- 1,520
- 3,583
You can open most files with notepad and the file should include the meta at the top of the file showing how the image was generated. In this particular case:
"full body HOT 18 year old adriana chechik with huge breast, abs, ((school girl)), (on campus) with Pamela Anderson, teacher, (mini skirt) hazel brown hair, jewelry, symmetry, digital painting
trending on artstation by Michael Garmash, (jeremy lipking), (photo), (detailed hands)
Negative prompt: ((((((male))), (((guy))), (((dude))), (((boy))), (((man)))))).
cartoon, ((disfigured)), ((bad art)), ((deformed)), ((poorly drawn)), ((ex-
tra limbs)), ((b&w)), weird colors, blurry
Steps: 38, Sampler: Euler a, CFG scale: 8, Seed: 1836225617, Size: 640x832, Model hash: 5a369d04a0, Model: pfg"
You must be registered to see the links
There are 2 ways to install Forge, One click if you don't understand how to use git all that well or you can just use git as usual.
You must be registered to see the links
Guide on how to install it using both methods.
You are welcome.
Last edited:
Thanks to Jimwalrus, I was able to get everything up and running and started experimenting late Thursday night into the wee hours like a demon. All images contain the necessary prompt data, but the idiotic thing is that I ended up closing out the Browser UI, and hence I cannot recall what Base Model I used, which sucks, as I wanted to refine these further. My 'Go Forward' approach is to screencap using the 'snipping tool' which is native to Windows and located under 'Windows Accessories'. I'll have to try and reverse engineer as I really like the BBW gals I generated.
I even practiced some technique refinement by directly plagiarizing Jim's amazing work from Page 1, to test the effectiveness of the "Restore Faces" functionality located on the 'Settings' tab within the UI. Thanks to everyone on this page for the help, I'm just a rookie, but this is a lot of fun. Fourth image is essentially how I envisioned Doris Day if she discovered the noon time Pizza Hut buffet (which is sorely missed, if you can forgive my editorializing). These are just directly out of stable diffusion, with native upscaling applied to the final image. I subscribe to an online image upscaler named Remini.ai, it does an amazing job at restoring the image quality of my collection of 35mm colour slides - for portraits, but it absolutely wrecks the detail of my vast collection of airliner and fighter aircraft slides.
I even practiced some technique refinement by directly plagiarizing Jim's amazing work from Page 1, to test the effectiveness of the "Restore Faces" functionality located on the 'Settings' tab within the UI. Thanks to everyone on this page for the help, I'm just a rookie, but this is a lot of fun. Fourth image is essentially how I envisioned Doris Day if she discovered the noon time Pizza Hut buffet (which is sorely missed, if you can forgive my editorializing). These are just directly out of stable diffusion, with native upscaling applied to the final image. I subscribe to an online image upscaler named Remini.ai, it does an amazing job at restoring the image quality of my collection of 35mm colour slides - for portraits, but it absolutely wrecks the detail of my vast collection of airliner and fighter aircraft slides.
- Jan 24, 2020
- 1,401
- 3,791
In order to see your settings etc and which ckpt model you used, simply load the image in png info and everything is right there including the ckpt model. You can send it to txt2img or img2img from png info and it will set everything for you. You only need to switch the ckpt manually. If you used postprocessing, you also need to enable it manually for some reason.
Thanks Mr. Fox - was able to locate the model and play around some more. Currently, really liking the revAnimated_v122EOL model. Just copied a few nice ones from civitai - these are just the direct output from Stable. Will look boss when run thru Remini.ai.
- Jan 24, 2020
- 1,401
- 3,791
There are several ways of upscaling in Stable Diffusion that most likely will give much better result than any external upscaler. It won't even come close.
Hiresfix in txt2img and SD Upscale in img2img tab. I have made posts about it in this thread. Seph might have linked to those posts on the 1st page.
*I see that you are using hiresfix. Go and get better upscalers. Ultrasharp and NMKD-Siax.
You must be registered to see the links
You must be registered to see the links
place them in Stable-Diffusion\Stable-Diffusion-WebUI-Forge\models\ESRGAN
If you don't have ESRGAN folder than simply create one.
Go to settings/upscaling and make sure " R-ESRGAN 4x+" is selected then press "apply settings". You also might need to reload the Ui.
Last edited:
- Jan 24, 2020
- 1,401
- 3,791
A tip is to not have hiresfix enabled while generating and looking for a good image (seed). Then when you find an image you like press the little star button under the image without doing anything else. It will then use the hiresfix settings and jump straight to the hires steps and not generate the basic image, this means you will retain the original image better.
Then send it to img2img, and go to script menu and select SD Upscale.
Now you will upscale the image one more time. Use a very low denoise setting, for a sharp image. You need to experiment a little to find the right denoise settings for you and your specific image as it can vary.
Try 0.1-0.2 as a start. The higher the denoise the more the image will change and vise versa, so you need to balance this.
The main resolution is now the tile size, and the multiplier in SD Upscale decides the final resolution based on the image you have in img2img.
This means that you can reduce the main resolution (tile size) and it will go quicker, only make sure the ratio is correct. Make sure to not use Face Restore or postprocessing while upscaling in img2img. It can create mask blend issues. It will show up as artifacts or pixelation around the head and hair.
Then send it to img2img, and go to script menu and select SD Upscale.
Now you will upscale the image one more time. Use a very low denoise setting, for a sharp image. You need to experiment a little to find the right denoise settings for you and your specific image as it can vary.
Try 0.1-0.2 as a start. The higher the denoise the more the image will change and vise versa, so you need to balance this.
The main resolution is now the tile size, and the multiplier in SD Upscale decides the final resolution based on the image you have in img2img.
This means that you can reduce the main resolution (tile size) and it will go quicker, only make sure the ratio is correct. Make sure to not use Face Restore or postprocessing while upscaling in img2img. It can create mask blend issues. It will show up as artifacts or pixelation around the head and hair.
Last edited:
- Jan 24, 2020
- 1,401
- 3,791
0.02 - 0.1 - 0.2 denoise strength. The difference is very very small and for this image it probably doesn't matter so much but the difference might be bigger in other use cases.
I'm just starting to learn to use ComfyUI. My main use will be animation. What are some good resources I could use to learn (or things to download)? I've loaded some workflows, but they always have nodes that I don't have, and there are usually no instructions on where to put the files.
- Mar 19, 2020
- 2,077
- 2,611
The first thing you should install with ComfyUI is the
You must be registered to see the links
. It's used to install custom nodes without you having to download those nodes manually via github and then add them to your CimfyUI installation yourself.Also: if a imported/downloaded workflow has red nodes, which means your installation is missing those nodes -- you can just use the ComfyUI-Manager to "Install Missing Custom Nodes".