must be config error
must be config error
Fbut what config? the other models works fine
Fi switch model and it runs fine
Fyou removed from prompting or training
Fi test that man class effect
Fclass images performed better
Fand if you mean training config..... no matter what I do. even if I replicate your videos 100%
Fi mean training configuration. can you try with training a man?
FYouTubeSECourses
Our Discord : https://discord.gg/HbqgGaZVmr. Grand Master tutorial for Textual Inversion / Text Embeddings. 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 Guide...
https://www.patreon.com/SECoursesPlaylist of Stable Diffusion Tutorials, Automatic1111 and Google Colab Guide...
FI can, but remember i trained this model perfectly fine 2-3 months ago
FYouTubeSECourses
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...
https://www.patreon.com/SECoursesPlaylist of #StableDif...
Fwhy one woman looks like that, and the other like this? xD
Fi can train your training images if you wish
Fon my computer
Fi'm using runpod, like 20 times or so
Fi'm gonna try the man and report back
Fit's weird because it doesn't look like overtrained or undertrained
Fit is possible to view metadata from a model like when you view it from a picture?
Flike.. base model, seed, steps/epochs, learning rate.... etc
Fis all that viewable on a model?
Fprobably not, but it's a good idea XD
Fi dont know if that is saved :d
Fi've see that you use "blue eyes" as the first negative prompt
Fi'm also always getting blue eyes
Fwhy do you think that happens?
Fbecause custom models are biased
Falso some tokens are biased
Flike handsome or photoshoot :d
Fi am going to sleep hopefully see you tomorrow
Fi spent like 8 hours to install new cpu and motherboard
F
Fi plan to compare all optimizers in kohya lora
AI removed from prompting.. I still use the class images.. but removing girl from training solved all the problems that I've had for months lol... I still don't know why it was such a specific problem with my one daughter.. probably because she var very hard for the model to determin her age.. my younger daughter who is 5 it had no problem with..
Fi thinking that this crappy regularization images that i'm using are deteriorating the results
Fthe woman class is re-trained with monsters
Fcould be nice find
Fcould be
Fi'm gonna try filewords + controlnet, someone told me that he's having a great result generating class images with controlnet depth map
Fso class images replicate the training images the most possible, but without the person face bias
Aif you buy multiple GPU's can use stack the VRAM to make training one model go faster?
like can you buy multiple 3090's in one computer and use them all on one A1111 image/model training?
like can you buy multiple 3090's in one computer and use them all on one A1111 image/model training?
Thow is amd looking for training right now? i want to buy a 7900xtx 
DHey guys what is the least amount of regularization images i need for artstyle and man?
DLike I am generating my own to train on top of a specific model so I can't wait for 300
DDaaamn i9-1300k did you also upgrade your gpu?
DMan I can relate to this so much it's actually spinning my mind now
KThat's awesome I can't wait to see your new video. And btw, in Kohya Dreambooth, what is the difference between "30 repeats x 10 epochs" and "10 repeats x 30 epochs"? Which option would produce better quality, or are they the same?
AI assume that its better for all of the images in the training data set to be in colour than to have 1 or 2 black and white ones?
Aany fix for this ?
AAre you creating your models from Hugginface, by checking this?
AIf not, then you are trying to trying to train a model before actually creating it. I had that happen to me yesterday.
ANo I using the pre-trained model, shown in this tutorial https://www.youtube.com/watch?v=kIyqAdd_i10
YouTubeSECourses
Our Discord : https://discord.gg/HbqgGaZVmr. This is the video where you will learn how to use Google Colab for Stable Diffusion. 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...
https://www.patreon.com/SECoursesPlaylist of Stable Diffusion Tutorials, #Automatic1111...