I did. Is there much quality difference between using Rank 3 config vs 4x_GPU_Rank_1_SLOW?
I did. Is there much quality difference between using Rank 3 config vs 4x_GPU_Rank_1_SLOW?
MIf not the only think I didn't try is training without captions
Fi think it could be all so every case has to be tested
Fyes small diff
Moh and I didn't try training T5 XXL. If there is no text in my training images would it still improve the training?
DAfter my failure I'm not so sure about "Apply T5 Attention Mask, and Train T5-XXL enabled" you are training the T5, but in inference you load a T5 model that doesn't know your token. forge o SwarmUI load the regular T5 model that was not trained. so I'm not so sure
Moh the t5 model should be trained as well for it to work properly?
D@Dr. Furkan Gözükara may be Apply T5 Attention Mask can help, but I don't understand how can Train T5-XXL helps, because on inference you load a T5 model that was not trained . forge or SwarmUI load the regular T5 model that was not trained
DIs the thing I don't understand, so really IDK now I'm training again with both options disabled, My question can apply to everyone, but in my case is different because I'm training flux-dev-de-distill so in my case is broking something. I'm having artifacts on the images with this on and also is not learning the concepts correctly
D@Dr. Furkan Gözükara Update on https://huggingface.co/nyanko7/flux-dev-de-distill I started new training - "T5 Attention Mask and T5-XXL both disabled" same lr 30 epochs now. I tested the lora checkpoint and all problems are fixed, is going great using regular flux-dev-fp8 for inference, I'm training tree people same class, same prompt changing just the name renders the correct subject, is undertrained but goes very well so far
A
Adid i do something wrong? why is there 17.6 million steps ??
FT5 XXL made minor diff that i didnt notice :d
Fyep you did indeed
Fsub folder
Fmore than 1 repeat
Aah it was at 40
Fyep
A
Ais that too many still
A2206 images
Fsub folders
VThe config could use an update as he made the same mistake as me.
Fi tested config today works as expected
VThe repeat was at 40 when I used it, is that not saved in the config?
Fit is read from folder
Fyou need to fix folder name
Fi explained this in details in tutorial
Fbut you are skipping :d
VI see. Lol.
F
![Dazzastrous [4090]](https://cdn.discordapp.com/avatars/871389071686135838/c4a5f4d81d9eeb18d856295d692e2c07.webp?size=40)
Dneed to be more vigilent?
Dhad to
Dlol
VSo here is what I suggest, as I watched some parts of all your tutorials, but not all (and sometimes the bookmarks are wrong). Add an image in the config file showing the repeat section and saying it needs to be 1.
VShould be labeled (Step 1: Read First!) or something similar.
Dyeh not the easiest to navigate through
Dso many videos
Dbit overwhelming for beginners
Fthat 40 repeat is not saved in config
Ftherefore i cant set it
Dwhy you get so many asking which video to watch
VI understand, hence the image I suggested.
Vadd the image with a screen shot of the repeats section. Then show a red arrow pointing at the box. With red text stating it should be 1.
V
Aah my bad, I just looked at configs and tried to figure out on the go 
DDoc how will you cope updating everything as more and more repos come out....
Dyou will have to be Dr Octopuss lol
V
D45 degree angle nice
