Another new day a new resumption

[Scarol]

I registered myself on civitai.com. And I saw the models, does the system that you have to generate your photos on the site work (which costs a lot of points each). I generated a few times but only vague faces came out, or do you have to save the photo of the model from that site and open it in Stable Diffusion? If I open an image and type in text, it keeps displaying the text "attribute error 'nonetype' object has no attribute 'lowvram'. What should I do now?

I used translate for this text, I can ony tape the bases english.

Here is a photo of an example.

Test 1.jpg

[cerberus]

I registered myself on civitai.com. And I saw the models, does the system that you have to generate your photos on the site work (which costs a lot of points each). I generated a few times but only vague faces came out, or do you have to save the photo of the model from that site and open it in Stable Diffusion? If I open an image and type in text, it keeps displaying the text "attribute error 'nonetype' object has no attribute 'lowvram'. What should I do now?

I used translate for this text, I can ony tape the bases english.

Here is a photo of an example.

Test 1.jpg

No you don't run anything on Civitai, you download the models etc from it, they're free. Pick a checkpoint model you want (AbsoluteReality and Dreamshaper are good ones) in the model page next to the "Create" button is a download button (Blue down pointing arrow and Tray), click this and the model file will be downloaded to your PC. Copy the downloaded file (should end in .safetensor) to your A1111/stable-diffusion-webui/models/Stable-diffusion folder.

In the A1111 browser page next to the model name (top left) is a blue button, this is the "list refersh" click this then click on the black triangle next to the model name and you should see a list of available models, which should now include the one you just copied in.

From your screen print you seem to be trying to use the face in the photo for the generated image. This isn't what Img2img is for, it's really for changing , fixing or upscaling images. What you seem to be trying to do requires an extension called ReActor and some other bits of software. I'd leave that for the time being and stick with the Txt2img tab while you get used to the basics.

[Scarol]

Ah okay, thanks Cerberus!

[Scarol]

Hi Cerberus, I did what you said but I have also to put the words --skip-torch-cuda-test after COMMANDLINE_ARGS in the document, otherwise the program won't start, is that normal?

I put the downloads in the folder and when I start the program and selects the file it won't show the black triangle and the models.
I downloaded 2 files but it only selects one and can't select the other.

I send you a print screen

[cerberus]

right coordinates... maybe 3 feet down?

you had me at Slow and Steady

the best way to fight claustrophobia is to take deep breaths...

Only 3 feet down! I don't think they've got room for deep breaths, too many collisions!

I'm back and right in time to keep on commenting before there are too many of those godies!

She's such a tease!
00036-1307123728.png

Nice one. Also hands are drawn really good!

Yep, didn't have to fix the hands on that one. I've had to do that a few times though.

"Um swimming movements don't seem to be working. What else you got?"
00673-1309608975.png

That ditch in this bog is quite nasty. Just put your arms right next to your body and then try to lift as much mud up as possible. Also try to bend and pump your legs simultaneously . Lets see if that helps, in any case i can search for a stick later....

If she follows those instructions you'll have to use the stick to find her!

No time for captions tonight, it's a bit late

00163-4185637605.png

00164-177948263.png

00187-2515408553.png

00188-1516852572.png

00233-217775115.png

[cerberus]

00352-570906685.png

00375-1171694089.png

00388-2940019761.png

00389-1096806960.png

00393-2213848895.png

[Scarol]

I succeeded to generate a photo with Stable Diffusion, now I will try the different modus.

Thanks!

[PA Jack]

Cerberus - your backgrounds continue to be great, very realistic and dramatic. Oh, and the models are pretty good, as well.

[cerberus]

I succeeded to generate a photo with Stable Diffusion, now I will try the different modus.

Thanks!

Sorry I missed your post last night, I was in a bit of a rush. It seems you worked out a solution anyway.

"Cuda" is the parallel processing platform on NVidia graphics cards, Stable Diffusion will use this if it's available and checks to see if it is. If you don't have a NVidia graphics card in your PC "--skip-torch-cuda-test" prevents SD from attempting to use cuda; so, yes in your case it's normal. I have a NVidia graphics card in my PC so I don't have this argument on my command line.

Sorry I wasn't that clear, the "black triangle" I mentioned is the bit circled in red on your screen print.

Test 2.jpg

I'd spend a bit of time playing with AbsoluteReality, trying different prompts and settings to get used to how things affect the images generated. CIvitai is useful here, on the model pages (e.g. Absolute Reality), if you click on one of the images, main or gallery (if you scroll down you get a gallery submitted by people using the model) there is some useful information displayed with the images. I've drawn a red box around it in the attached screen print. This information includes the prompt, negative prompt, the CFG Scale - labelled "Guidance", the sampler and scheduler, number of steps etc. Copying and experimenting with some of these settings can help focus the experimentation.

Screenshot_3.jpg

[MadMax359]

not too late to rescue damsels (eventually!)