Thanks a lot, appreciate the help 🙂
Thanks a lot, appreciate the help 
Fyep looks like slower than before
Ffor some reason
F
Fi will wait till 100 steps
Foh i didnt do rtx 5090 bat let me do that
Fit could be my gpu throttling right now
Fover 90 celcius
F
Fyep slower than before
F
Flet me test same on rtx 5090
GYup, those look same s/it as mine, really glad you tested it, was going crazy thinking I was doing something wrong
Fyep no error at your side
Fblock swapping slower than before
Gah ok, thanks a lot for double checking it. Guess there's nothing we can do then right other than hope the dev fixes it at some point?
Fprobably. it is really weird how it is broken somehow
GHey I'm new around here and have very little coding knowledge, but I'm getting these error messages when installing ultimate image processing V15 on massedcompute. Would appreciate any help 
Thanks
Thanks
SI followed this post. I got YOLO/SAM2 Detected Images, but I don't get the cropped images..
SI got this image, but not the cropped one.
Ffirst one not error
Fbut second one is
Flet me test
Fyou selected this right @g7ants80
Fit doesnt crop
Fwhat you need exactly?
Fi mean it crops to aspect ratio not the target
Fif you tell me what you need i may code it
Fhello sorry for this error fixed with v17
FPatreon
Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon
SIt's working now, I am using version 11, and it works with version 11. sir! We can do it easily with other software like birme.org, xnConvert or other online tools.
NFyes
GThank you! I'll see if it works now
GGetting this now when I start training
GAnyone have any idea what GPU on MssedCompute is the closest equivalent to RTX 5090's 3s/it for finetuning? (single only, not duals). Feel the need for speed :p
Fit tells you the error
Fdid you set your training images folder?
Fh100 faster
Fa100 can be close
GThanks Doc
DI am currently following and running the training from: https://www.patreon.com/posts/kohya-flux-fine-112099700
Once the training is finished, can the output file be used as a LoRA data file in ComfyUI using the Flux model?
Sorry if my question is confusing, as I am still a beginner.
Thank you!
Once the training is finished, can the output file be used as a LoRA data file in ComfyUI using the Flux model?
Sorry if my question is confusing, as I am still a beginner.
Thank you!
Patreon
Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon
DI am a beginner and would like to ask:
If I train a model using FLUX Full Fine-Tuning or DreamBooth, can the resulting trained file be used as a LoRA model in a ComfyUI workflow?
Sorry if my question is basic, I am still learning. Thank you!
If I train a model using FLUX Full Fine-Tuning or DreamBooth, can the resulting trained file be used as a LoRA model in a ComfyUI workflow?
Sorry if my question is basic, I am still learning. Thank you!
JHi, thanks for your excellent tutorials and articles.
I'm having an issue with runpods and LoRA. I completed some tests locally and know have started over with runpods. When I
Am I missing something else?
Disregard, loading different base configs seemed to fix it.
start trainingLoRA type must be set to 'Flux1', 'Flux1 OFT' or 'LyCORIS' if Flux1 checkbox is checked.Disregard, loading different base configs seemed to fix it.
F100%
Bfor 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