There will be support for 1.1?
There will be support for 1.1?
C1.1 is flux pro. flux pro, all versions, are api only
CI personally doubt it, I don't see the reason for them releasing it
LI'm doing local with 4090
Ci'm going to suggest you go to the technical support channel on the stable diffusion discord and look for CS1o, and ask him about that. that question comes up in there all the time
Dthank you\
LMaybe there is better checkpoints instead of training flux? Like people not fine tune the model to do more realistic? Like sdxl models
CIf you are doing the fine tune just be aware when it saves checkpoints they are huge, if you are wanting to just tinker around for fun I think a lora is a decent first option. Just get like 15 pics that are half decent and see how it goes
LI already did lora and it did very well
CIf you are wanting a better checkpoint for the people realism (eg skin) part I think you will likely find that your own training/lora will already address that enough as is probably
LHmm it interesting
CDepends what your goal is of course. Personally I have just trained of dev and have been very happy with the results as far as skin texture etc but im just messing around for fun with friends family so not super picky. The only time I trained something other than dev was when I trained a lora of our old dog using the checkpoint i trained of my wife as using two loras was giving me junk results
LWell I trained on my pics and lot of the pictures doing smooth skins i wondered if it because of of my data set maybe some of the pics has smooth skin filter
LI trained on 33 pics
Cthe AI is only going to learn what it's taught ...
Cit's a computer, it's not about to guess
IYo
ISo can we fine tune the whole model now?
Cno
IOh then I don't get what furkan posted this weekend xD
LLet's do our own research and create better model!
Ccool - you'll need a high powered data center, a couple million dollars, and a staff of research scientsts.
Lsounds easy
IMoney 
Cif it was easy, everyone and their dog would be putting out killer models
Bhey @Dr. Furkan Gözükara Are there any additional benefits to using the DreamBooth/extract method for fine-tuning Flux compared to directly training a LoRA, besides achieving higher image quality? For example, does it offer better resistance to overfitting, more flexibility, or differences in training time?
Byup, making your own models requires tons of capital for the staff and compute required to pull it off. It's why there only a few startups out there with their own models and lots of AI companies using API
Cjust putting the data set together, correctly, for a base model, is quite a task that requires expertise - never mind training the thing
BYeah dataset preparing and gathering requires a ton of time and resources
Cyes, and you're not going to just go scrape google for images - that's why most of them use the LAION library
Cbillions
Cof images
Cat least 2 billion if you're just making a tiny model
BI’m planning to build a SaaS product based on some ComfyUI workflows I’ve created, using Flux as the main image generation model. I’ve spoken with the team at Fal.ai, and they confirmed I can use it commercially through their platform, as long as it runs through their API due to their arrangement with Blackforestlabs. They also mentioned they can host ComfyUI and that it’s possible to use ComfyUI as a backend for a SaaS product, as long as all models used in the workflow (like upscalers, facial detection, etc.) are licensed for commercial use. I’m still deciding whether to stick with ComfyUI or build a custom pipeline based on my workflows.
PHey Dr. Furka, my friend is training a flux character lora (not realistic) with default settings of ai-toolkit and I try training with kohya and your config rank 3, and he seems to get really good results, but I'm having a problem getting the proper body shape and some details properly.
It seems like no matter what parameters I try, it feels undertrained.
tried 1 repeat with 200 epochs (20 images) lr0.00005
tried 20 repeat with 10 epochs (20 images) lr0.0001
tried 150 repeat with 1 epoch, (save every 1000 steps) tried the output of 2000, 3000 ,4000, and unless I make the prompts very long and detailed, I'm getting something that resembles the character and clothes, but the body shape and face mostly gets random every seed and it won't really capture the character.
Even if I put a lot of details in the prompt, it kinda works, but not really.
And yet, my friend's first attempt with default settings on ai-toolkit, which has much less parameters to change, gets awesome results and accuracy, and it's still very flexible while keeping the essence of the character pretty well.
Any idea what I'm doing wrong?
It seems like no matter what parameters I try, it feels undertrained.
tried 1 repeat with 200 epochs (20 images) lr0.00005
tried 20 repeat with 10 epochs (20 images) lr0.0001
tried 150 repeat with 1 epoch, (save every 1000 steps) tried the output of 2000, 3000 ,4000, and unless I make the prompts very long and detailed, I'm getting something that resembles the character and clothes, but the body shape and face mostly gets random every seed and it won't really capture the character.
Even if I put a lot of details in the prompt, it kinda works, but not really.
And yet, my friend's first attempt with default settings on ai-toolkit, which has much less parameters to change, gets awesome results and accuracy, and it's still very flexible while keeping the essence of the character pretty well.
Any idea what I'm doing wrong?
Pforgot to mention, we trained the same character
his captions are mostly automated with roughly manual fixes
my captions were automated at first, then highly reconstructed manually in a consistent captioning way
his captions are mostly automated with roughly manual fixes
my captions were automated at first, then highly reconstructed manually in a consistent captioning way
Psame dataset too
PRight now I'm trying to replicate his decent+ result of ai-toolkit, but with kohya_ss, just to have a solid starting point, but I can't manage to achieve that.
BAnyone here trained their own upscaler model before?
SAny recommendations on how to setup A1111 on AWS?
SOr Forge
Fyou dont need caption if your dataset is consisten enough
Fenable shared vram
Finitial loading uses VRAM
Fi trained on my rtx 3060 yesterday perfectly training
Fwell some people want to share and keep lora
Fonly benefit it is
Fotherwise original checkpoint best
Fdont use captions
Fwhen you are training a single concept
Fif your concept is consisten, captions reduces likeliness
