Software Engineering Courses (SECourses)•17mo ago

Can confirm batch size 1 is fastest (in my case anyways)

sergio_gonzo•9/1/24, 1:59 PM

Dr Furkan the loras FLUX could work in any checkpoint or must be in specific ones?

Afroman4peace•9/1/24, 2:12 PM

hi all

Ssergio_gonzo Dr Furkan the loras FLUX could work in any checkpoint or must be in specific one...

ThereiverOP•9/1/24, 2:14 PM

I believe flux dev loras only work with flux dev, and flux schnell loras for flux schnell. AFAIK they're not compatible with SD 1.5, SDXL etc

sergio_gonzo•9/1/24, 2:24 PM

thank you man, but i saw there is a variant in size of flux dev i must see sometinh like that, if have f16 maybe?

Ssergio_gonzo Dr Furkan the loras FLUX could work in any checkpoint or must be in specific one...

dxqb•9/1/24, 2:29 PM

I tried a character lora on the "New Reality" FLUX preAlpha checkpoint. results weren't good

dxqb•9/1/24, 2:30 PM

so probably same as SDXL. some work better than others, but generally Loras don't translate well between checkpoints

Afroman4peace•9/1/24, 2:35 PM

I am a checkpoint trainer myself (Colossus Project XL) yes some checkpoints are trained in very unique ways that may can result in bad trainable models

AAfroman4peace I am a checkpoint trainer myself (Colossus Project XL) yes some checkpoints are ...

dxqb•9/1/24, 2:37 PM

can I ask you an unrelated questions?
what do you guys do that somehow screws up the possibility to train embeddings on your models? I haven't tested Colossus, but on all other finetuned SDXL checkpoints like RealVis, Juggernaut, ... you cannot train embeddings, only on SDXL base. Same with Pony: you can train embeddings on PonyDiffusionXL, but not on any derivates like PonyRealism

Afroman4peace•9/1/24, 2:41 PM

most of all many guys in the community loves to copy stuff. At the beginning of SDXL I was the first that combined the refiner model with the normal model. I think the first one was my Hephaistos model... not long after that nearly everyone used it.

Afroman4peace•9/1/24, 2:42 PM

Then there are some sdxl models that have seperated from the original so much that its nearly something completely different than sdxl base

dxqb•9/1/24, 2:43 PM

I don't mean that it's different or that the embedding doesn't translate well between checkpoints. it's impossible to create an embedding on finetuned SDXL models

Afroman4peace•9/1/24, 2:46 PM

I haven't trained an embedding on my models. There was simply no need to do it. I may will test this

Afroman4peace•9/1/24, 2:51 PM

@Dr. Furkan Gözükara I am here to say thanks for your tutorial and the kohya_ss settings. I will use this to train new Flux models. I have just a 4070ti with 12GB so this will be a great help for me

dsienra•9/1/24, 2:52 PM

@Dr. Furkan Gözükara Dr do you know what is "Split QKV" is good or bad for quality and what it does?

virginia•9/1/24, 3:06 PM

@Dr. Furkan Gözükara ：（ my images from swarmUI is blurry, i do not know what is wrong, i mean flux

ThereiverOP•9/1/24, 3:21 PM

What resolution were your training images?

ThereiverOP•9/1/24, 3:21 PM

And are your training images bad quality?

Vvirginia @Dr. Furkan Gözükara ：（ my images from swarmUI is blurry, i do not know what is ...

dxqb•9/1/24, 5:32 PM

lower CFG to 1

James K•9/1/24, 8:05 PM

Flux can run on M3 18GB Mac ?

JJames K Flux can run on M3 18GB Mac ?

Maxivi•9/1/24, 8:37 PM

sell your mac, and get a desktop, and dual boot linux or use windows with wsl

MMaxivi sell your mac, and get a desktop, and dual boot linux or use windows with wsl

James K•9/1/24, 8:38 PM

I am addicted.

James K•9/1/24, 8:38 PM

Now Changing to Windows is very hard.

JJames K Now Changing to Windows is very hard.

Maxivi•9/1/24, 8:38 PM

not really i use both, windows is mad easy

MMaxivi not really i use both, windows is mad easy

James K•9/1/24, 8:39 PM

Idk. I already paid apple tax

JJames K Idk. I already paid apple tax

Maxivi•9/1/24, 8:39 PM

yeah you can still sell it and get a decent amount back for a m3 macbook

Maxivi•9/1/24, 8:39 PM

i will sell mine soon, collecting dust

MMaxivi i will sell mine soon, collecting dust

James K•9/1/24, 8:55 PM

I don't do AI tasks regularly. & All my work is on my Mac. So it doesn't make sense for me to move out

sergio_gonzo•9/1/24, 9:04 PM

i make my first lora and the result is superb, any other lora that i make before no was this level of good until a check onetrainer there was my first sucessfull models with the likeness and still no this level of good, what could be the evolution of flux? wich model can make a step forward like this ?

GGomezBro even batch size 4 works on A6000 bros I tested it today

Furkan Gözükara SECourses•9/1/24, 9:29 PM

yes but no speed increase

Furkan Gözükara SECourses•9/1/24, 9:29 PM

after 2

Furkan Gözükara SECourses•9/1/24, 9:29 PM

pretty useless - you just lose quality

TTOPUI When will this latest JSON configuration file be updated?

Furkan Gözükara SECourses•9/1/24, 9:29 PM

hopefully today sorry for delay

Furkan Gözükara SECourses•9/1/24, 9:29 PM

i came back from mountain house fully

Furkan Gözükara SECourses•9/1/24, 9:29 PM

finally 500 mbit

Furkan Gözükara SECourses•9/1/24, 9:30 PM

actually looks like i got upgrade for some reason

TThereiver Can confirm batch size 1 is fastest (in my case anyways)

Furkan Gözükara SECourses•9/1/24, 9:30 PM

batch size 1 isnt fastest - you need to calculate effective speed

Furkan Gözükara SECourses•9/1/24, 9:31 PM

but batch size 1 is best quality

FFurkan Gözükara SECourses pretty useless - you just lose quality

GomezBro•9/1/24, 9:31 PM

yup the lora was useless

GomezBro•9/1/24, 9:31 PM

bad quality

Ssergio_gonzo Dr Furkan the loras FLUX could work in any checkpoint or must be in specific one...

Furkan Gözükara SECourses•9/1/24, 9:31 PM

i didnt test quantized

Furkan Gözükara SECourses•9/1/24, 9:31 PM

but should work on fp8 and fp16

GomezBro•9/1/24, 9:31 PM

you had mentioned that that 200 EPOCH is for training people. What is the epoch to train a style? and why

AAfroman4peace hi all

Furkan Gözükara SECourses•9/1/24, 9:32 PM

hi welcome

GGomezBro you had mentioned that that 200 EPOCH is for training people. What is the epoc...

Furkan Gözükara SECourses•9/1/24, 9:33 PM

it is a good question. actually there is no hard limit. but i try to limit it 200 epoch so not very aggresive learning rate or not very low learning rate either

Furkan Gözükara SECourses•9/1/24, 9:33 PM

it is like trying to balance

Furkan Gözükara SECourses•9/1/24, 9:33 PM

probably you can go 1000 epoch with lower LR and get better quality but it just becomes infeasable timing

Furkan Gözükara SECourses•9/1/24, 9:33 PM

so for style if you have too many images

Furkan Gözükara SECourses•9/1/24, 9:33 PM

go as much as you can endure

Furkan Gözükara SECourses•9/1/24, 9:34 PM

and compare every 1 checkpoint

Furkan Gözükara SECourses•9/1/24, 9:34 PM

lets say you have 1000 images

Can confirm batch size 1 is fastest (in my case anyways)

Similar Threads

Similar Threads

Similar Threads