what is the message?
what is the message?
Fif you dont have at least 12 gb vram add this to command line arguments
F--ckptfix
Glet me give that a try
GSuccess
Zi'm a generate 2400 images of woman, then set my num class instances of image to 50 and see how that does me
GI just want to understand that, I am running a 3070Ti and I have dreambooth running now after adding the --ckptfix arag. Do I continue as normal or Do I need to enable Lora?
ZWhat is the relationship between classification dataset directory and Class images per instance image
Fwhich one is best to use in thumbnail
Fcontinue as normal after that. that is for generate training model
Gperfect.
GSecond row third column
GI can use the regularization images you have supplied here right ?
Fuse same class prompt
GI don't quite understand the filewords like instance token or class token or instance prompt or class prompts. I will continue watching but evaerything up until this point has been fairly straight forward
JThis is a wiki guid but I do not understand it either hope it helps you: https://github.com/d8ahazard/sd_dreambooth_extension/wiki/Class-Explained
GitHub
Contribute to d8ahazard/sd_dreambooth_extension development by creating an account on GitHub.
Fexplained in this video
Fwait it
GYa, I get the idea around concepts but what throws me off is that I have no idea what fileworlds, traning prompts and sample prompts do . I will continue with the video but it's not the 1st time watching. I feel it's a difficult concept to grasp
Jwhich video exactly, then I would also look at it maybe we can understand it together
GYouTubeSECourses - Software Engineering Courses
Our Discord : https://discord.gg/HbqgGaZVmr. The most advanced tutorial of Stable Diffusion Dreambooth Training. 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 Stable Diffusion Tutorials, Automatic1111 and Google Colab ...
https://www.patreon.com/SECoursesPlaylist of Stable Diffusion Tutorials, Automatic1111 and Google Colab ...
GSo I am currently adding the filewords
FYes that one is really comprehensive
G
FI don't put them in filewords tab
FUsing in training prompts better
FEasier to understand also
FExplained in New video hopefully will be published in 1 hour
GI look forward to it
AHello everyone! glad to find this treasure trove. Was wondering if there are any resources to learn dreambooth for objects such as a lamp? Thanks in advance.
JI must confess it was a bit too fast for me.
I will have to make a kind of pre flight checklist so that I can use at first because it is so much at once.
I will have to make a kind of pre flight checklist so that I can use at first because it is so much at once.
JI don't know anything about this, but I have seen that in dreambooth there is "Traing Wizard (Object/Style)" under Concepts:
Aappreciate thanks a lot. will try it out n report
JIf you understand Classen and Filwords you are welcome to return the favor. 
Gstuck on this as well XD
ASure thing 
THey guys. I just ran a two gpu in one system unit for Sd rendering or training. I was wondering how did Colab ran multiple gpus? I cant still figure it out on local sd. Anyone doing like what im doing here? 
JOut of interest, are these two systems or one PC with two graphics? Vertical / Horizontal upgrade?
J@Arshy
@Garreth210
Would you have time and interest that we sit together tonight, research and exchange information. (Voice)
Maybe we can help each other.
@Garreth210
Would you have time and interest that we sit together tonight, research and exchange information. (Voice)
Maybe we can help each other.
GSure, I don't mind
GThis is my current setup
G
GI am just trying things at this point
AI’m GMT+5.5. Not sure if that works for u guys. Also later in the week would be better for me. But I had be interested
G@Furkan Gözükara SECourses Question:
If I use the reference images you have supplied on gDrive, Do I still need to set the class image generation ?
If I use the reference images you have supplied on gDrive, Do I still need to set the class image generation ?
G
JAs far as I understand, no, if you start from the standard model and have a similar training targets. Otherwise it is recommended. (I think the reference images should correspond to the trainigs gole.
Gah
Gokay
Gso then I should be able to hit train and shit should theoretically start working