Software Engineering Courses (SECourses)•6mo ago

Hey, just wanted to throw this here FYI in case you're looking for STAR alternatives within your APP

Hey, just wanted to throw this here FYI in case you're looking for STAR alternatives within your APP, you can check if you want, they claim to be better than STAR, however, I haven't been able to test it out.
https://github.com/yjsunnn/DLoRAL?tab=readme-ov-file

GitHub

GitHub - yjsunnn/DLoRAL: Official Implementation of “One-Step Dif...

Official Implementation of “One-Step Diffusion for Detail-Rich and Temporally Consistent Video Super-Resolution” - yjsunnn/DLoRAL

HHenry Hello! I am trying to get the kaggle notebook to run "kohya-sdxl-lora-training-o...

Furkan Gözükara SECourses•7/7/25, 10:09 PM

yes we had to update a lot

HHenry Also, I can seem to edit the GPUs or the folder directories to point to training...

Furkan Gözükara SECourses•7/7/25, 10:10 PM

yes you have to edit them

CCryptozow.x Hey, just wanted to throw this here FYI in case you're looking for STAR alternat...

Furkan Gözükara SECourses•7/7/25, 10:10 PM

i just checked

Furkan Gözükara SECourses•7/7/25, 10:10 PM

looks weak

Henry•7/7/25, 10:10 PM

I got it to work now!

HHenry I got it to work now!

Furkan Gözükara SECourses•7/7/25, 10:11 PM

great

underdog•7/9/25, 12:25 AM

Hello,

I'm using the configuration you have been using for 48 gb ram. I rented an H100 GPU with 20 image dataset but the images generated after training does not even match with my training data. Where do you think I might be going wrong? all the parameters are loaded from 15 image dataset config file. Can you please advise?

Uunderdog Hello, I'm using the configuration you have been using for 48 gb ram. I rented ...

Furkan Gözükara SECourses•7/9/25, 12:29 AM

which config you used which zip file

Furkan Gözükara SECourses•7/9/25, 12:29 AM

please more info

FFurkan Gözükara SECourses which config you used which zip file

underdog•7/9/25, 12:30 AM

I used 48GB_GPU_28200MB_6.3_second_it_Tier_1.json

Uunderdog I used 48GB_GPU_28200MB_6.3_second_it_Tier_1.json

Furkan Gözükara SECourses•7/9/25, 12:31 AM

ok it is a dreambooth config. did you load into dreambooth tab?

Furkan Gözükara SECourses•7/9/25, 12:31 AM

properly did set training images folder?

Furkan Gözükara SECourses•7/9/25, 12:32 AM

also we have an example dataset

Furkan Gözükara SECourses•7/9/25, 12:32 AM

https://www.patreon.com/posts/114972274

Patreon

Perfect Quality Example Model Training Images Dataset - Can Be Used...

Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon

underdog•7/9/25, 12:33 AM

yes I'm loading them in the dream booth tab. I also am loading the training images properly but just to be sure is there a way for me to know if the training images are being loaded properly or not?

underdog•7/9/25, 12:35 AM

I can see the folder being created after clicking prepare training data

Uunderdog I can see the folder being created after clicking prepare training data

Furkan Gözükara SECourses•7/9/25, 12:36 AM

ok this is accurate

Furkan Gözükara SECourses•7/9/25, 12:36 AM

you can look at the cmd logs

Furkan Gözükara SECourses•7/9/25, 12:36 AM

to see

Furkan Gözükara SECourses•7/9/25, 12:36 AM

if you copy paste here i can understand

underdog•7/9/25, 12:39 AM

I get this when I press prepare training data

/kaggle/working/kohya_ss/training_data/img/1_ohwx
man...
00:37:48-667704 INFO Regularization images directory is missing... not
copying regularisation images...
00:37:48-670635 INFO Done creating kohya_ss training folder structure at

and when I start training, you can see them in the message.txt file.

Thanks a lot for your help.

message.txt546.13KB

underdog•7/9/25, 12:51 AM

I tried other images including mine but didn't work, so I tried taking a few screenshots from the 15 image dataset of yours and at the 1000th step the sample image I get is not even close to the training images. I will try with Dwayne Johnson images now and give you an update.

gFspxwNfsQP3re7IZvx8hk0A5lXeyp8hgGgdEKqffwAxUyyrgpMhgAAAABJRU5ErkJggg.png

Uunderdog I get this when I press prepare training data /kaggle/working/kohya_ss/trainin...

Furkan Gözükara SECourses•7/9/25, 12:52 AM

last time i tested

Furkan Gözükara SECourses•7/9/25, 12:52 AM

kaggle was not able to train flux

Furkan Gözükara SECourses•7/9/25, 12:53 AM

"/kaggle/working/kohya_ss/traini
ng_data/img/1_ohwx man"

underdog•7/9/25, 12:53 AM

oh? so what do I use? I'm using jupyter notebook through vast.ai

Uunderdog oh? so what do I use? I'm using jupyter notebook through vast.ai

Furkan Gözükara SECourses•7/9/25, 12:57 AM

it says kaggle

Furkan Gözükara SECourses•7/9/25, 12:57 AM

is that accurate?

Furkan Gözükara SECourses•7/9/25, 12:57 AM

well vast ai python version i see 3.12

Furkan Gözükara SECourses•7/9/25, 12:57 AM

vast ai template could be wrong i dont know

Furkan Gözükara SECourses•7/9/25, 12:58 AM

i test on runpod and massed compute

underdog•7/9/25, 12:59 AM

I'm not sure either, because the interface I'm using is jupyter lab but the folders being created says kaggle.
Let me try using massed compute and I will update you with the results I get. Thanks for your advise.

Uunderdog I'm not sure either, because the interface I'm using is jupyter lab but the fold...

Furkan Gözükara SECourses•7/9/25, 1:01 AM

sure you are welcome

Furkan Gözükara SECourses•7/9/25, 1:01 AM

also as a test you can use our example dataset

Furkan Gözükara SECourses•7/9/25, 1:01 AM

but with flux

Furkan Gözükara SECourses•7/9/25, 1:01 AM

it is very easy to get maximum resemblance

underdog•7/9/25, 1:04 AM

I'm using flux dev for finetuning. Where can I find your example dataset? I couldnt find them in the finetuing post. Do you mind sharing the link?

Uunderdog I'm using flux dev for finetuning. Where can I find your example dataset? I coul...

Furkan Gözükara SECourses•7/9/25, 1:08 AM

https://www.patreon.com/posts/114972274

Patreon

Perfect Quality Example Model Training Images Dataset - Can Be Used...

Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon

underdog•7/9/25, 1:09 AM

oh sorry I thought you were talking about a different dataset. I'm going to use this dataset and try it out using massed compute now.

Uunderdog oh sorry I thought you were talking about a different dataset. I'm going to use ...

Furkan Gözükara SECourses•7/9/25, 1:20 AM

great

FFurkan Gözükara SECourses great

underdog•7/9/25, 5:02 AM

update -Dwayne johnson training works fine. I get good results around the 3000th step. I will try again for other characters to see if matches with the training data or not. Thanks for your help. I have one more question, is there a way to programatically select the best checkpoint instead of manually checking the checkpoints to see which one is the best?

Uunderdog update -Dwayne johnson training works fine. I get good results around the 3000th...

underdog•7/9/25, 7:59 AM

Another question - In the dataset Dwayne johnson does not have hair but when I generate images a few images for Dwayne Johnson hair is being added, do you know why and how can we keep it real? I can add no hair in the negative prompt but is there a different way to keep it generic without hardcoding.

Elli_Oaken•7/9/25, 8:38 AM

HI! I have a question about lora-training and captioning. A friend of mine suggest that when you caption your object your not supposed to mention the object in the image at all, only everything around in like a semantic mask and where the object-tag works as the caption for the lora-object. On youtube-tutorials on the other hand everyone I've seen says that you should in detail concistently describe the object and only mention shortly the rest of the image.

What would you guys say is the case here?

Uunderdog update -Dwayne johnson training works fine. I get good results around the 3000th...

Furkan Gözükara SECourses•7/9/25, 1:54 PM

sadly not

Furkan Gözükara SECourses•7/9/25, 1:54 PM

you need to evaluate yourself

Uunderdog Another question - In the dataset Dwayne johnson does not have hair but when I g...

Furkan Gözükara SECourses•7/9/25, 1:54 PM

yes it is about prompt

Furkan Gözükara SECourses•7/9/25, 1:54 PM

just add bald :d