- Jun 4, 2022
- 155
- 69
Well still, on frist load of SD all first checkpoints load faster, than the next ones. Like you load SD, 1st checkpoint 0 sec, 2nd 3 seconds, 3rd 20 sec, 4th 50... etc
Yes, if you load the previous checkpoint it will load fast, but all the new will be loading slower and slower, while I have 0 VRM set to save the hash of checkpoints. I think this setting for hash tend to save hash of checkpoints forever, even if you close SD on the next load you won't need to load hash again. Maybe that's why people have SD .cache file over 100 GB or smth
- Mar 17, 2021
- 308
- 1,053
For who is interested i released a
Also released some other, check my profile on civitai.
You must be registered to see the links
.You don't have permission to view the spoiler content.
Log in or register now.
- Dec 23, 2018
- 519
- 2,151
Not sure if you were messing about with inpainting here but your entire prompt was
parameters
Rachel Green and Monica Geller
Negative prompt: blurry
which wasn't really that helpful. I just inpainted rachel first and changed the prompt to 'rachel green, jennifer aniston, friends' then swapped over to 'courtney cox, monica geller, friends'
Rachel's head looks weird to me though
- Jan 4, 2022
- 233
- 869
yes it is , other is ControlNet
trouble here is for posing hands,legs,bots. Image is in comics style.
- Nov 12, 2017
- 407
- 3,182
- Apr 21, 2019
- 13
- 15
Just started using SD two days ago.I am having a blast with it.Here are some images i love!
- Dec 21, 2020
- 1,531
- 3,618
The files contain ComfyUI prompts.
You don't have permission to view the spoiler content.
Log in or register now.
Last edited:
- Jan 24, 2020
- 1,401
- 3,794
"If the saved image file size is above the limit, or its either width or height are above the limit, save a downscaled copy as JPGA small question.
For some reason txt2img-grids saves 2 files .jpeg and .png (two dublicated grids), I don't need .jpeg but can't find where to disable it in settings, it shows only png
File size limit for the above option, MB"
Either unselect this box or increase the size limit.
- Jan 24, 2020
- 1,401
- 3,794
- Apr 21, 2019
- 13
- 15
- Jan 24, 2020
- 1,401
- 3,794
I think she has gotten some shaving cream on her shoulder, for some reason.. Does Asian girls have hairy arms?
- Mar 24, 2019
- 1,010
- 3,418
Hi, need some help here please. I tryed to reproduce an image on but it doesnt work and I dont understand why.
I get this Lora :
I'm trying to make this one to check if it works good :
And this is what I have after (I think) copying every parameter:
The author said he used anytrollV4 to train his Lora, for what I understood, the name was some kind of joke, and the real model was anythingv4 which I used. I just wish to know if the problem come from the model or if I'm missing something that I do not understand yet.
I get this Lora :
You must be registered to see the links
I'm trying to make this one to check if it works good :
You don't have permission to view the spoiler content.
Log in or register now.
And this is what I have after (I think) copying every parameter:
You don't have permission to view the spoiler content.
Log in or register now.
The author said he used anytrollV4 to train his Lora, for what I understood, the name was some kind of joke, and the real model was anythingv4 which I used. I just wish to know if the problem come from the model or if I'm missing something that I do not understand yet.
- Mar 24, 2019
- 1,010
- 3,418
Top pic model hash : 5cab7be36e
Bottom pic model hash : d5f8393fea
You have 2 different versions of model here, so it's not perfect copy as you are claiming. Different models were trained on different data, so nartually for the same parameters they yeild different end result.
Last edited:
- Mar 24, 2019
- 1,010
- 3,418
Not that I know of any easy way, but how do you know it's anything model to begin with ?
I tryed some search for AnytrollV4, and couldnt find it, some people said it was a fake update of anythingV3, some other said it was a variant of an orangemix one. Since I'm new in stable diffusion I'm not sure I understood it well.