ok thanks, I think I understand it better now
ok thanks, I think I understand it better now
Fyep because multi gpu increases batch size but keep same step count
Fso 1 step you actually do 2
DMaybe that was the difference then as I trained the LoRAs with AdamW but the dreambooth finetune with Adafactor (I could not fit AdamW in my gpu or a rented 48GB one)
Do you think this could be the difference?
I still believe finetuning should be better so I’m just confused about why LoRA yields better results for me.
Do you think this could be the difference?
I still believe finetuning should be better so I’m just confused about why LoRA yields better results for me.
Fhard to know possibly a mistake somewhere 
AHey there!
Anyone out here was deploying flux lora training as an API?
Would be happy to chat or be grateful if you can share something useful on how to deploy to runpod/other compute providers.
P.S. If you can share some useful presets for training person lora - beer on me haha!
Anyone out here was deploying flux lora training as an API?
Would be happy to chat or be grateful if you can share something useful on how to deploy to runpod/other compute providers.
P.S. If you can share some useful presets for training person lora - beer on me haha!
HDoes training lora for Noobai v-pred require different settings?
I'm using kohya_ss gui (dev branch that is up to date with the latest script).
Someone somewhere said that it needs
I'm using kohya_ss gui (dev branch that is up to date with the latest script).
Someone somewhere said that it needs
--zero_terminal_snr
Fwhat is Noobai v-pred
Wrecently when people train lora with kohya and they also train clip text encode together, they claim that lora train together with clip text encode give much better result. I wonder your latest lora training script json file have these clip text encode trained with lora or not ?
WIt a finetuning of Illustrious model, another heavy modified training from SDXL model.
HNoobAI-XL v-pred uses v-pred instead of eps like other SDXL models which makes it behave a bit different.
DI'll keep testing and let you know. In general the deambooth finetune does preserve small details betetr (logos, brake caliper, etc), but the LoRA follows the car's shape a lot better. I need to test it further with a contolnet and depth map (taken from the 3D model) to see if this car body issue can be resolved that way. And maybe as a final test I can send a 80GB+ GPU to do a AdamW dreambooth training run.
HAfter a test I can say that normal SDXL lora training settings +
--zero_terminal_snr
Tare you sure fine tune checkpoint is not just undertrained? How many images do you have in the trainign dataset and how many training steps did you do for lora and fine-tuning?
TThe thing is - lora training is just much faster in terms of steps/resemblance compared to full checkpoint, but checkpoint should yeild better generalization
Tthis chat is full of friendly people sharing their knowledge of training flux, for starters just follow basic tutorials of Dr. Furkan on the subject you are interested, they cover 99% of your questions
DGood point, thanks for your suggestion. The dataset has 45 images. LoRA produces good results at around 6000 steps. For dreambooth I am trying around 9 - 10K steps. Do you think I should train longer still?
Tyes, continue training if you don't see model sanity degradation, you can get better results compared to lora at 18-20k steps or more, if your use case require perfect resemblance at the cost of lower model sanity
Tbut if your use case do not benefit much from improved generalization this can be just too costly for similar results
DGreat thanks I will give that a try. Costs and speed are not a priority, obtaining the best product accuracy / quality is, so I will give this a go by training a lot longer and compare the epochs in a grid. Cheers
Fwe already train clip
Four all lora configs have it 
DWould it help to also train the text encoder (at 50% unet LR for example) and use T5 attention mask for dreambooth training just as we do with LoRA training? Has anyone seen any benefits?
WCompare with train checkpoint follow your config, if I extract lora from trained checkpoint, is the extracted lora quality better ?
GAnnybody finetuned a style using image data set of a particular city district? I will start It over a previous finetuned person using dreambooth
Fkohya not supporting
DWhat do you mean? Both parameters can be applied in the koyha UI (dreambooth tab) and the training runs without issues. Do you mean it will cause problems with the final trained models?
Fkohya doesnt train clip large or t5 when doing fine tuning
Ftry you will see same speed and vram
DAh I see, thanks for clarifying
Fyes i told to add kohya but he still didnt add 
DInteresting to test when it does get implemented at some point.
K@Furkan Gözükara SECourses Hi sir, we trained the Lora model using a quality Tier 1 config file with 48GB of VRAM.
We generated an AI character with photo grid, cropped, flipped, and upscaled the images. We created a dataset of about 8 images to train the Lora model.
While generating images from Lora, the close-ups have a really good face, but in wide shots, the face becomes pixelated.
Additionally, the face does not closely resemble the training dataset.
I’m attaching the images as well.
brown background image is generated
white background image is dataset
We generated an AI character with photo grid, cropped, flipped, and upscaled the images. We created a dataset of about 8 images to train the Lora model.
While generating images from Lora, the close-ups have a really good face, but in wide shots, the face becomes pixelated.
Additionally, the face does not closely resemble the training dataset.
I’m attaching the images as well.
brown background image is generated
white background image is dataset
TI think the dataset is not good, the same background and clothes in all the photos? Probably the problem is in the dataset
TBut I don't know, sometimes these things happen 
THow many photos did you train?
K8 images , actually
KWhich is the best way to train a AI generated character then
KAnd which is the best way to create a dataset for AI generated character?, I tried multiple things, the consistent is not so good
KCan someone help me in this it will be really helpfull
�try creating consistent images with pulid first and then train on them
Kthanks nitin , for your suggestion , will try it ,and show the results here
Fi recommend 2 stage
Ffirst train
Fgenerate better images
Fthen collect more images
Fand train again
KShould I train the first trained Lora with a new dataset using more steps, or should I train from scratch using the images generated by the trained Lora as a dataset?
Ffrom scractch with new dataset
Kcool