hello all. love the videos!
hello all. love the videos!
Zgot a strange error during training "Returning result: Exception training model: Inference tensors cannot be saved for backward. To work around you can make a clone to get a normal tensor and use it in autograd.
"
"
RI'm trying to install Stable Diffusion on my computer, I have followed all the steps carefully, but when I try to run it by clicking "webui-user.bat" the result is like this, and not as shown in the tutorial. anyone know what mistake i did? can anyone help me how to fix this?
Qhmm nvm i did. let me look whats diffrent with mine
Qit looks like your commit hash is none you try launch.py
Qyour missing the hash. it sohuld be Commit hash: a9fed7c364061ae6efb37f797b6b522cb3cf7aa2 or something
SHey i have a question I trying to train a checkpoint with dreambooth using the Realistic Vision V1.3 as the base model. What i am attempting to do is train the following yoga pose dataset https://www.kaggle.com/datasets/niharika41298/yoga-poses-dataset into realistic vision so the model will be able to generate people in various yoga poses. would I just download all the images and then label each of them with the selected pose? Any pointers or is there any resources available for me 2 look at for a task like this. Thank you
Qyou should look into controlnet posing with stable diff.
Qthere is a openpose editor
Q
SI can do it with controlnet no problem, but the application i am making I want to be able to generate these on demand
Qhmm sorry then i dont know how i can help you still learning alot with training my own models. if you found it out. i would love to hear it
Shoping the Dr. @Furkan Gözükara SECourses will have some insight on this. I have been watching his videos all day long today
Fyou cant. you have to kill initial one and start another terminal. still you can connect via 3000 port after that
Fjust delete entire folder do a fresh install . i explained in the beginning of this video : https://youtu.be/tXaQAkOgezQ
YouTubeSECourses - Software Engineering Courses
Discord : https://bit.ly/SECoursesDiscord. New fantastic style transfer feature via T2I-Adapter added to the #ControlNet extension. 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 #StableDiffusion Tutorials, #Automatic11...
https://www.patreon.com/SECoursesPlaylist of #StableDiffusion Tutorials, #Automatic11...
Fif you cant change background than you need to modify your training dataset and make different backgrounds
Fthat means model memorized them
Fhi welcome
Fit is easy with invoke ai
Fhello welcome and thanks
Fhappens when generating preview or after that?
Fhi welcome
Fwhen you type python to new cmd what do you get
Fprobably you need reguler dreambooth training
Fbut it usually not work on custom models especially cooked ones
G@Furkan Gözükara SECourses this is what mine looks like /etc/sysctl.conf
G
GThe face ins near perfect but like I mentioned, None of my prompts work no matter the length pos or neg prompts
Fdid you save different checkpoints
Fthat is what matter most
GI did, however the moment I go a checkpoint down then my face does not show anymore
Gso I will retrain the model but I'm not sure if it's over training or what it is.
I have tried different checkpoints but I go toward the last checkpoints and my face looks good with the blue eyes or I got one checkpoint to early and my face simply does not show up. THere is no middle ground that I could find
I have tried different checkpoints but I go toward the last checkpoints and my face looks good with the blue eyes or I got one checkpoint to early and my face simply does not show up. THere is no middle ground that I could find
Gbut I can comfortably set up runpod and everything needed to start training
Fhow many per epoch do you save a checkpoint
Falso how is your training dataset that matters a lot
GTraining dataset, different shirts, different backgrounds
Fhow many per epoch do you save
Falso have you tried different prompts?
GI have tried MANY MANY MANY propts and I think it was 10 per epoch
Fin your example | didnt work
GCorrect. It did not work but it was the last thing I tried before deleting my pod
Fi see
Fif you are our patreon supporter i can setup training for you via connecting your computer
GThat sounds like a good idea. I might look into that.
Scan anyone please help me with loha training T_T
Fwhat help you need
SIdk which parameters i should be using
Shere is what i have rn
SLoHaBasicSettings.json2.12KB
Fye cant help with that. it requires extensive experiment and i didnt make yet
Fhow much vram you have
