so you are training on colab and using ckpt in your pc right?
so you are training on colab and using ckpt in your pc right?
ZDr Furkan, everytime I go on reddit, i see your name. You're the man
Sthats what i hope to do
but it didnt work out, and somehow i managed to generate images in colab now...nevertheless, the model i trained doesnt seems to be as great as those one your demonstrated on video. Guess the sample images is not good enough (maybe)
but it didnt work out, and somehow i managed to generate images in colab now...nevertheless, the model i trained doesnt seems to be as great as those one your demonstrated on video. Guess the sample images is not good enough (maybe)
Fthank you so much for comment
Fsample images matter a lot
Zi must have screwed up my filewords or something cause this is what my model trained to
Zwas training a woman, but i put the person name under instance token and woman under class token
Zgonna watch your video in a little bit cause i forgot how to do that lol
B@Furkan Gözükara SECourses What you recommend me to do in steps to train a specific subject (person)? Like, for e.x first make a textual inversion, after that a lora, then generate different class images and finally do dreambooth, that could be a nice step by step practice?
BThe goal is to create a style and then add the specific subjects to it
FIs there a way to set up the training so that the learning rate changes as it progresses?
Fstart with 2e-6 and after some time go to 1e-6 and so on
FI read that you can do something like this with linear scheduler and setting the warm up steps
Fbut not sure what value to put on warm up
SI am so grateful, thank you. I did as you displayed in the gif and got the error below. I went to the file and unclicked read only (it was a - instead of a checkmark) and it made me apply to all subfolders. I ran it again and got the same error. Please Don't kick me out!
Zcan i use torch adamw or lion optimizer with lora?
SLooks like I fixed it! Now to learn how to train with it. Thank you!
CI have a question about saving a merge
COn colab
SHello guys, i have a problem, can anyone help me please? I install dreambooth on stable diffusion locally, and when i'm trying to ai learning, im getting weird pictures and they are not look's like as imput images at all
MAre you using filewords?
MIf not then your prompt is in the wrong place
MWatch this https://youtu.be/Bdl-jWR3Ukc
YouTubeSECourses - 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 ...
MIt’s a bit out dated but the basic concepts are there.
SThanks for the answer! Really appreciate that, i'll watch
MBecause you are training a style you want to make sure you use a class token as style or something like that.
MIf you are not familiar with Dreambooth I would recommend training a subject first like yourself just to get an idea of what is going on first. Style is hard. Dr Furkan Gözükara is still making a video on that.
SThanks for explanation, i've watched a video and now i have a new problem
Si'm trying to setup this program like for 4 days really
F-Hi! I´m a digital artist that has been working on VFX and 3D animation for 20 years,-I´m not a programmer, never dedicated time to learn code. -I'm totally aware of this revolution that is happening exponentially. -I have been studying and learning AI for the last couple of months, (mostly Stable Diffusion Auto1111 WebUI - Deforum - Dreambooth) and at this point I have some questions about the tools available.1- I have seen that Midjurney ( at this point ) is absolutely superior compared to Stable Diffusion but I'm not sure if that is because I´m using an older model: v1-52- Are there plans to upgrade SD models soon? I know that there are some models around the internet that you can download but is there a formal project to upgrade the standard model soon? 3- As a noob I'm not sure if one of the pros of SD Auto1111 WebUI is that you can create your own models and produce animations using extensions like Deforum. Is that the main reason many people are using it? Sacrificing quality, I mean....4- Noob question: can you load the Midjurney model into Auto1111 WebUI? 5- Why Midjurney results are SO SUPERIOR when you compare them to SD Auto1111 WebUIThanks!!!
Fcurrently i am recording this video
Fwhat title do you suggest it?
Fteached a style and myself
Fgreat
Fyes possible but i dont find it is better
Fyou are using incorrectly the filewords
Fin this video I have explained
FYouTubeSECourses - Software Engineering Courses
Our Discord : https://discord.gg/HbqgGaZVmr. Newest update of DreamBooth extension of Automatic1111 brought huge quality and success improvement. 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...
https://www.patreon.com/SECoursesPlaylist of Stable Diffusion Tutorials...
Fhello. i suggest to do fresh install and create a new training. alternatively if you are our patreon supporter i can connect your pc and setup everything for you if you wish
F1: midjourney is fine tuned model
Fcurrently i am producing a video to teach style
Fit can consistently generate outputs like this
F
F
F2: yes i think sd version 3 coming soon
F3 : i didnt understand the question part
F4 : no you cant
F5 : it is super fine tuned model that is why 
SCan you please send to me a link of the patreon site (your)?
FPatreon
Become a patron of SECourses today: Get access to exclusive content and experiences on the world’s largest membership platform for artists and creators.