for what reason do we extract lora after dreambooth?
for what reason do we extract lora after dreambooth?
DHello, I’m still a beginner and learning through SECourses, so I apologize if my question is unclear.
I successfully trained a character using Flux Full Fine Tuning (or DreamBooth), and I tested the result in ComfyUI as a UNet model (without using flux1.Dev). The results were very realistic and successful.
Now I’m wondering:
I successfully trained a character using Flux Full Fine Tuning (or DreamBooth), and I tested the result in ComfyUI as a UNet model (without using flux1.Dev). The results were very realistic and successful.
Now I’m wondering:
- Can this trained model be used as a replacement for Flux Kontent?
- Or, can I use Flux Kontent and combine it with my trained model as a LoRA?
Fif you need lora
Fotherwise you dont need and i dont recommned dreambooth works better
F1 : nope different model
F2 : nope sadly
Fbut i didnt test lora on kontext yet so cant say for 100%
�I want to fine tune on a specific quality type of images of women. Not 1 character, but I want the fine tune to be able to generate woman with same quality as training data. Say I have pics of 5 to 10 different women of multiple shapes, hair, body, ethnicity etc. How many pics would be needed for a high quality finetune? Ideally it should be able to generate women not in the training set, I.e. an Arab woman, but with same quality as training pics. Unsure how to proceed, only done fine tune with 1 character so far. I guess captions should be needed here.
Fyou may try bigger fine tunes people published with good prompting
Fotherwise you are looking for rather general fine tune
Fso you need more images the better and use good captions
HHello! I am trying to get the kaggle notebook to run "kohya-sdxl-lora-training-on-a-free-kaggle-notebook_v28", but the moment I import it just starts downloading a ton of dependencies automatically. I did not see this in the tutorial. What is the expected behavior? Thanks in advance
HAlso, I can seem to edit the GPUs or the folder directories to point to training images
CHey, 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
https://github.com/yjsunnn/DLoRAL?tab=readme-ov-file
GitHub
Official Implementation of “One-Step Diffusion for Detail-Rich and Temporally Consistent Video Super-Resolution” - yjsunnn/DLoRAL
Fyes we had to update a lot
Fyes you have to edit them
Fi just checked
Flooks weak
HI got it to work now!
Fgreat
UHello,
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?
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?
Fwhich config you used which zip file
Fplease more info
UI used 48GB_GPU_28200MB_6.3_second_it_Tier_1.json
Fok it is a dreambooth config. did you load into dreambooth tab?
Fproperly did set training images folder?
Falso we have an example dataset
FPatreon
Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon
Uyes 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?
UI can see the folder being created after clicking prepare training data
Fok this is accurate
Fyou can look at the cmd logs
Fto see
Fif you copy paste here i can understand
UI 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.
/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
UI 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.
Flast time i tested
Fkaggle was not able to train flux
F"/kaggle/working/kohya_ss/traini
ng_data/img/1_ohwx man"
ng_data/img/1_ohwx man"
Uoh? so what do I use? I'm using jupyter notebook through vast.ai
Fit says kaggle
Fis that accurate?
Fwell vast ai python version i see 3.12
Fvast ai template could be wrong i dont know
Fi test on runpod and massed compute
UI'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.
Let me try using massed compute and I will update you with the results I get. Thanks for your advise.
Fsure you are welcome
Falso as a test you can use our example dataset
Fbut with flux
Fit is very easy to get maximum resemblance
UI'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?