Awesome video. Thanks for sharing it!
Awesome video. Thanks for sharing it!
FYouTubeSECourses
In this video I have intensively compared RTX 5090 speed on FLUX DEV, FLUX Fill, SD 3.5 Large, SD 3.5 Medium, Stable Diffusion XL (SDXL) and Stable Diffusion 1.5 (SD 1.5) models. For each benchmark, I have compared RTX 5090 against RTX 3090 TI so we see the speed improvement. Moreover, I have tested FP8 vs 16-bit precision for FLUX and SD 3.5 La...
Mi used vpn and it worked, but its not accepting any of my cards, ill try to get a USA prepaid card today
A@Dr. Furkan Gözükara or anyone here tried new Pika Labs Addon feature? I need feedback and opinions
Fi didnt use Pika Labs yet
Mbtw is this gold the highest tier there?
Fyes. what you need? if you need private consultation let me know via private message
Mi am not looking for something specific like private consultation, community support would be enough
Fyes gold is great
Fworking on hunyan
Ais it worth it tho?
AImage2vid seems more promising for research hours
V@julius @Dr. Furkan Gözükara
Ah, okay. My problem is that I have too many neutral-looking, generic images in my training set. I included them thinking more images = better results, but now I’m noticing some issues.
Face Inpainting Issue
I realized that during image generation, the face initially appears to be smiling in the preview, but by the end, it looks more neutral. It seems like the Face Inpainter is neutralizing the expression.
Questions:
Does the image generation process actually have two stages?
Step 1: Normal image generation
Step 2: Face inpainting
How does the Face Inpainter (YOLO) work?
Does it pick a face from my training images, or is it using something else?
Can I adjust the Face Inpainter?
For example, can I tell it which face image to use as a reference?
Can I turn off the Face Inpainter for testing?
I'd like to see how the results change without it.
Training Set Strategy
I’m planning to retrain the model with a new set of training images—this time, focusing on more expressive (smiling, emotional) faces instead of too many neutral or redundant ones. Right now, I have to generate tons of images just to get the expression I want, which is not ideal.
How can I quickly test if a training image set is good?
I don’t want to spend 45+ hours training again only to realize my dataset was bad.
Can I use a lower-tier config for a quick test?
For example, could I train with "6GB_GPU_5400MB_14.1_second_it_Tier_3_512px.json" on a 48GB A6000 GPU (Massed Compute) just to get a fast idea of what expressions the model generates?
Or would the results be completely different compared to using a high-end config like "48GB_GPU_28200MB_6.3_second_it_Tier_1.json"?
Thanks!!!
Ah, okay. My problem is that I have too many neutral-looking, generic images in my training set. I included them thinking more images = better results, but now I’m noticing some issues.
Face Inpainting Issue
I realized that during image generation, the face initially appears to be smiling in the preview, but by the end, it looks more neutral. It seems like the Face Inpainter is neutralizing the expression.
Questions:
Does the image generation process actually have two stages?Step 1: Normal image generation
Step 2: Face inpainting
How does the Face Inpainter (YOLO) work?Does it pick a face from my training images, or is it using something else?
Can I adjust the Face Inpainter?For example, can I tell it which face image to use as a reference?
Can I turn off the Face Inpainter for testing?I'd like to see how the results change without it.
Training Set Strategy
I’m planning to retrain the model with a new set of training images—this time, focusing on more expressive (smiling, emotional) faces instead of too many neutral or redundant ones. Right now, I have to generate tons of images just to get the expression I want, which is not ideal.
How can I quickly test if a training image set is good?I don’t want to spend 45+ hours training again only to realize my dataset was bad.
Can I use a lower-tier config for a quick test?For example, could I train with "6GB_GPU_5400MB_14.1_second_it_Tier_3_512px.json" on a 48GB A6000 GPU (Massed Compute) just to get a fast idea of what expressions the model generates?
Or would the results be completely different compared to using a high-end config like "48GB_GPU_28200MB_6.3_second_it_Tier_1.json"?
Thanks!!!
F@VioletAnt please watch the entire tutorial. also all tier 1 is equal quality
Fand example dataset shared here
Ftake a look
FPatreon
Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon
F1 : yes
F2 : it mask and inpaint
F3 : what do you mean
F4 : yes
Fworking on hunyuan
Fwow hunyuan is slow
Fanyone got good hunyuan video prompts to test?
Fteacache difference
Fpretty minimal
F
Mhow can i load a json config file in kohya?
Fevery tutorial showing it 
Fi recommend watch this
FYouTubeSECourses
If you want to train FLUX with maximum possible quality, this is the tutorial looking for. In this comprehensive tutorial, you will learn how to install Kohya GUI and use it to fully Fine-Tune / DreamBooth FLUX model. After that how to use SwarmUI to compare generated checkpoints / models and find the very best one to generate most amazing image...
A@Dr. Furkan Gözükara
ASwarm stopped showing me how the image is made up
AWhat happened? yesterday it was good I didn't change anything
Mweird, i watched this video but i dont recall imporing json configs, lemme check it again
Mbtw i got this error, and the loss is becoming a NaN, would this affect the training?
Fyes Nan means not working
F
Fconfig load
Fno idea
Fmake sure to load dreambooth configs into dreambooth lora into lora tab
ABtw what's TeaCache?
Fincreases speed
AIt says Install button bellow Preferred D type
Fi am testing today
Ashould I implement? I'm on kaggle
M... i generated a 10k steps lora this morning like this, no wonder it screwed up 
Mhow can i fix this error?
Fi would say get a fresh config into lora tab
Falso verify your images
Fany values is fine except nan