I have questions about Flux Lora training especially about prompting for characters:
I have questions about Flux Lora training especially about prompting for characters:
SDo you prompt at all?
Do you use tags or natural language?
Do you use tags or natural language?
SWhat do you describe in the prompt?
Everything that is important about the character or everything that is not important about the character?
Or everything in the training picture?
Everything that is important about the character or everything that is not important about the character?
Or everything in the training picture?
DI use "Charactername class and the normal prompt" or "Charactername class, prompt"
Din the prompt put all you want, the clothing the setting location what is doing, etc
Sare you using kohya? if so how do you feel these boxes:
Insance prompt: diego man,
and class prmpt: man .
Insance prompt: diego man,
and class prmpt: man .
DInsance prompt: Diego
Dclass prmpt: man
Dthat's for training
Dthen promt like I said
DDr. Furkan have a great list of prompts
FHey, @Dr. Furkan Gözükara I think 256 images is a lot, can I make 64 for a Dreamooth?
DYes, you can
Fthanks
F@druhl
Fi shared test prompts check them out
Fand i dont caption
Fonly
Fohwx man
Fyou see even 15 images works :d
Fkohya is broken right now - newest version
FGitHub
Urgent fix needed for SD3 branch @kohya-ss this error is after latest @bmaltais update this config worked yesterday INFO prepare accelerator flux_train.py:177 accelerator device: cuda Saving traini...
AHey guys, I'm looking to do a big finetune on FLUX1. Dev, not just a training for a single character, more like a big model finetune.
The aim is to get more realistic looking images of people. I know it was possible for SD 1.5 as I currently work with finetuned version of SD 1.5 that is making more realistic images than FLUX.
I have a few thousand high quality and well labelled images, and can get a lot more if that is required, so my main question is:
Is this possible to do with flux yet ? If so, can anyone point me in the direction of githubs or resources to find more info on how I can do this
The aim is to get more realistic looking images of people. I know it was possible for SD 1.5 as I currently work with finetuned version of SD 1.5 that is making more realistic images than FLUX.
I have a few thousand high quality and well labelled images, and can get a lot more if that is required, so my main question is:
Is this possible to do with flux yet ? If so, can anyone point me in the direction of githubs or resources to find more info on how I can do this
Fi tried with 10k images
Fcurrently flux get broken
Fyou can test my checkpoint here
F5 epoch : FLUX_Photography_2e_06-000005.safetensors
F10k images
AInteresting, I'll try it out now. Do you have any idea whats causing it to break ?
Fi will test converting checkpoints into fp8
Fbeing distilled model
A
Fi am waitin them to become more mature
Fnext will research them
Asounds like a plan
Asorry for all the questions but do you have an estimate for how long you think that will be
Fsadly i dont know but i hope be fast
Fmy next is recording fine tuning video
Fcurrently testing FP8 conversiton of models
Ftesting
Fwow almost no quality loss
F
Awow
Awhat FP is the original
F
F
Foriginal is fp16
DWhat got broken exactly?