Software Engineering Courses (SECourses)•3y ago

Every tutorial I've watched says nothing about providing class images. They always just seem to gen

Every tutorial I've watched says nothing about providing class images. They always just seem to generate them?

RewireOP•3/25/23, 9:20 PM

Picture of a man

RewireOP•3/25/23, 9:20 PM

For example

idmon•3/25/23, 9:20 PM

You can generate them yourself and save them for future trainings

idmon•3/25/23, 9:21 PM

Store them in a folder "man", "woman", "..." all the classes you frequently train

idmon•3/25/23, 9:21 PM

and link back to that

idmon•3/25/23, 9:21 PM

saves time

RewireOP•3/25/23, 9:21 PM

Interesting

RewireOP•3/25/23, 9:21 PM

I mean after the training I can generate pictures of my face just fine but attaching them to something else like a superhero costume doesn't do anything anytime I insert my instance prompt it just generates a picture of my face it won't ever do anything extra beyond that.

RewireOP•3/25/23, 9:23 PM

Can I use my own preprocessed images with clip descriptions to act as a class?

RewireOP•3/25/23, 9:23 PM

Or will that skew the training data.

idmon•3/25/23, 9:23 PM

Yes I do that myself actually

idmon•3/25/23, 9:24 PM

I generate them on the same model I train on (important detail) on sampler: DDIM (default diffusion method for training) on 30 - 50 steps

RewireOP•3/25/23, 9:24 PM

Interesting. Any luck with adding additional prompts though like making yourself into a superhero or making yourself into like iron Man or something.

RewireOP•3/25/23, 9:25 PM

I just can't seem to get past the step of generating my own face it does that just fine but I can't attach it to anything.

RewireOP•3/25/23, 9:25 PM

Wondering if there's an issue with my workflow.

idmon•3/25/23, 9:25 PM

I usually try to generate class images that look very similair to my subject. I try to write detailed prompts and do a little bit of "class bleeding"

idmon•3/25/23, 9:26 PM

It can help with converging your model faster using class images

RewireOP•3/25/23, 9:26 PM

Hmm

idmon•3/25/23, 9:26 PM

RewireOP•3/25/23, 9:26 PM

I'll try that when I get home.

idmon•3/25/23, 9:26 PM

instead of doing just "man"

idmon•3/25/23, 9:27 PM

I went for specific key features

idmon•3/25/23, 9:27 PM

still experimenting, but so far helped with this specific dataset

RewireOP•3/25/23, 9:27 PM

Yes I see now. I guess I'll have to fiddle with it more when I get back home.

RewireOP•3/25/23, 9:28 PM

I'm just trying to get myself looking like I'm sitting at a high-end bar in a business suit drinking a gin and tonic or something. But it just generates pictures of my face after I do the dream booth training which is effectively useless.

RewireOP•3/25/23, 9:28 PM

I feel like my prompts are pretty detailed but whenever I add my personal instance prompt tag for my training it just generates pictures of my face and doesn't do any outpainting or additional content.

idmon•3/25/23, 9:29 PM

Then it's overtrained, the UNet is overcooked

idmon•3/25/23, 9:29 PM

How many steps did you train?

idmon•3/25/23, 9:29 PM

How many images did you had?

RewireOP•3/25/23, 9:31 PM

Oh well then that makes sense yeah I mean whenever I follow the tutorials online for training or use the built in "wizard" I end up getting to like 12,000 training steps

RewireOP•3/25/23, 9:31 PM

My data set was 40 images.

RewireOP•3/25/23, 9:48 PM

I guess I'm used to old machine learning where you would have to train for thousands and thousands of steps before getting good results.

RewireOP•3/25/23, 9:48 PM

Things are better these days

felixsanz•3/25/23, 9:50 PM

generate several ckpts in the process and you can compare them

RRewire I guess I'm used to old machine learning where you would have to train for thous...

idmon•3/25/23, 10:37 PM

mostly thanks to Google open sourcing all these great breakthroughs in optimizers and learning schedulers

idmon•3/25/23, 10:37 PM

we need less compute to get the same training results

idmon•3/25/23, 10:38 PM

also Facebook now with xformers you can speed up a lot + Nvidia Cudnn installed

Ggenialgenteel @Furkan Gözükara SECourses Hello, I am new here, and I am trying out training a ...

Furkan Gözükara SECourses•3/25/23, 10:41 PM

welcome. sorry for late reply

Ggenialgenteel @Furkan Gözükara SECourses Hello, I am new here, and I am trying out training a ...

Furkan Gözükara SECourses•3/25/23, 10:41 PM

you can ignore warnings just follow the flow.

Furkan Gözükara SECourses•3/25/23, 10:46 PM

stupid discord cant ban the spammer

CCypo @Furkan Gözükara SECourses

Furkan Gözükara SECourses•3/25/23, 10:47 PM

trying to ban discord doing nothing

Bbrenkhof Click to see attachment

Furkan Gözükara SECourses•3/25/23, 10:48 PM

hi . you have followed this one? https://youtu.be/oI0GEnrySS0

YouTubeSECourses - Software Engineering Courses

How To Install New DreamBooth Extension On RunPod - Automatic1111 W...

RunPod: https://bit.ly/RunPodIO. Discord : https://bit.ly/SECoursesDiscord. How to install and start using #DreamBooth extension on #RunPod SD Web UI template. If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on

https://www.patreon.com/SECourses

DreamBooth extension: ht...

How To Install New DreamBooth Extension On RunPod - Automatic1111 W...

LLongshao Hey guys, just downloaded realistic vision to my laptop. How do I get it to work...

Furkan Gözükara SECourses•3/25/23, 10:48 PM

here have full tutorial : https://youtu.be/QN1vdGhjcRc

YouTubeSECourses - Software Engineering Courses

Ultimate RunPod Tutorial For Stable Diffusion - Automatic1111 - Dat...

Sign up RunPod: https://bit.ly/RunPodIO. Our Discord : https://discord.gg/HbqgGaZVmr. This is the Grand Master tutorial for running Stable Diffusion via Web UI on RunPod cloud services. If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on

https://www.patreon.com/SECourses...

Ultimate RunPod Tutorial For Stable Diffusion - Automatic1111 - Dat...

Iidmon Because if you don't provide it... it will generate it based on this value

Furkan Gözükara SECourses•3/25/23, 10:49 PM

i have 2400 photo of man pre generated publicly available description of this video : https://youtu.be/Tb4IYIYm4os

YouTubeSECourses - Software Engineering Courses

Automatic1111 Stable Diffusion DreamBooth Guide: Optimal Classifica...

Sign up RunPod: https://bit.ly/RunPodIO. Our Discord : https://discord.gg/HbqgGaZVmr. New best training settings for DreamBooth training in Automatic1111 Web UI. If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on

https://www.patreon.com/SECourses

Playlist of #StableDif...

Automatic1111 Stable Diffusion DreamBooth Guide: Optimal Classifica...

RRewire I cant get a good model out of dreambooth, and training takes hours on my 3090. ...

Furkan Gözükara SECourses•3/25/23, 10:49 PM

must be error. watch this video extremely helpful : https://youtu.be/sRdtVanSRl4

YouTubeSECourses - Software Engineering Courses

Epic Web UI DreamBooth Update - New Best Settings - 10 Stable Diffu...

RunPod: https://bit.ly/RunPodIO. Discord : https://bit.ly/SECoursesDiscord. Web UI DreamBooth got epic update and we tested all new features to find the best settings (10). If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on

https://www.patreon.com/SECourses

Playlist of...