i mean big beards, mine is a full foot long probably more
i mean big beards, mine is a full foot long probably more
EHello, need some help on copying some SD files from one pod to another, watched the tutorial on youtube but it errors out. Not sure if this is the right channel to ask, if not, apologies in advance.
Zyeah nothing i seem to do will get this womans face properly
Fupdated script you can redownload
Fi think yes why not
Tprobably my dataset then, I have a feeling the transparency of beards can cause issue
Zdoc my wife is going to divorce me if i don't make a good model of her
Zno pressure
II have entered discord, because I was a teacher, and they are the best tutorials for the use and installation of stable diffusion that I have seen of all the ones on youtube, not only for the explanations of how to do the things, but there is an addition where You explain, for example, in dreambooth the parameters for which they are used in detail, then you also apply different resolutions and compare the results.
It is let's say a very scientific methodology, congratulations
It is let's say a very scientific methodology, congratulations
Zis having 200 images in my dataset using 5 class reg images the same as using 10 images in my dataset using 5 class reg images
Zlike how do i get it to learn every intricate detail of someones face
Fthank you so much
Fyes I am also a teacher. Assistant professor in a private university in Turkey
Fnew video almost ready. waiting youtube to synch subtitles hopefully
ZI did a search on stable diffusion, saw an article on linkedin, guess whose it was
Zis the only point of the num class instancers per image or regularization images is to preserve the class?
Zso if it overfits, if i type woman, it will only show the woman in my dataset, and no other women?
Zcan you explain doc?
Z@Furkan Gözükara SECourses
Fnot only that
Fit also prevents overtraining
Fif you train very long, even when you dont type anything you get your trained subject
Fbecause unet becomes cooked
Zovertraining is when the images start looking warped?
Fusually
TI have a question, when you use the "ohwx man" and the <lora> to generate an image, are you still increasing the weights using ctrl+up e.g. (ohwx man:1.5) or are you keeping it at 1.1?
2nd question; Do we need both [the lora and the ohwx man] everytime we want to call our model?
2nd question; Do we need both [the lora and the ohwx man] everytime we want to call our model?
Fi usually test this because sometimes 1.1 or 1.2 works better
Fso you need to test
F2nd : yes
Tthank you!
I am seeing great results with 1.3 but gets worse higher than that
can you increase the weights on the <lora>? what effect would that have?
I am seeing great results with 1.3 but gets worse higher than that
can you increase the weights on the <lora>? what effect would that have?
Factually i didnt test increased weight of lora but you can test easily with sr prompt
TAlright, thanks again!
Zwhat's the best way to impose someones face using controlnet?
Zi like codeformer but it seems to change the face a little too much, should i lower the weight? or try gfpgan
IAhhh! and a good teacher. Because good teachers explain things as if they were learning the first time.
It is not easy to explain things as if you were explaining them to yourself, when you are already an expert in the field.
But for students it is the best way for them to learn.
And in your tutorials you can see this way of teaching
It is not easy to explain things as if you were explaining them to yourself, when you are already an expert in the field.
But for students it is the best way for them to learn.
And in your tutorials you can see this way of teaching
FThanks agreed
Ro/
Rany alternatives to kohya for LoRa training?
Ror anyone able to help me with my 4090 just getting 3.5it/s?
Fkohya is slow
Fdreambooth extension of web ui much better
Falso better quality
R
Fhave you seen my videos?
Ri think i just watched one 
Ron kohya
Rbut can i do LoRa in dreambooth extension of automatic1111?
Ralso is 16it/s in line with what to expect from a single 4090 in automatic1111?
A@Furkan Gözükara SECourses hello. I just watched your last video. Amazing stuff. One question. You said you trained The model based of The settings in that video but then you said you generated The images using realistic vision. I dont understand how that works.. Or did you train The images on realistic, because in The video you use The standard model..
Fyes you can do
F16 it is low