Software Engineering Courses (SECourses)•16mo ago

dmn model learnt even my broken teeth haah 😄

dmn model learnt even my broken teeth haah

Furkan Gözükara SECoursesOP•9/11/24, 9:35 AM

i will now retrain batch size 1 with up to 30 epochs

Furkan Gözükara SECoursesOP•9/11/24, 9:36 AM

after 30 epochs it became cooked

Furkan Gözükara SECoursesOP•9/11/24, 9:36 AM

maybe maximum 40 epochs

Furkan Gözükara SECoursesOP•9/11/24, 9:37 AM

it learnt my body shape too perfectly

Furkan Gözükara SECoursesOP•9/11/24, 9:37 AM

Furkan Gözükara SECoursesOP•9/11/24, 10:17 AM

T5 attention mask + T5 XXL speed is 4.52 second / it on L40S - training everything

daroth•9/11/24, 10:19 AM

for generation is Euler Simple still the best bet?

Ddaroth for generation is Euler Simple still the best bet?

Furkan Gözükara SECoursesOP•9/11/24, 10:20 AM

iPNDM is best i think

daroth•9/11/24, 10:21 AM

simple?

daroth•9/11/24, 10:21 AM

scheduler I mean

Ddaroth scheduler I mean

Furkan Gözükara SECoursesOP•9/11/24, 10:27 AM

normal - default - didnt change it actually

Mattheus Chediak•9/11/24, 11:43 AM

FLUX LoRA Training Simplified: From Zero to Hero with Kohya SS GUI (8GB GPU, Windows) Tutorial Guide
do we have a video for this but using runpod ?

Mattheus Chediak•9/11/24, 11:50 AM

i found already, nevermind

MMattheus Chediak i found already, nevermind

Furkan Gözükara SECoursesOP•9/11/24, 11:59 AM

ye it has both runpod and massed compute here : https://youtu.be/-uhL2nW7Ddw

YouTubeSECourses

Blazing Fast & Ultra Cheap FLUX LoRA Training on Massed Compute & R...

Unlock the power of FLUX LoRA training, even if you're short on GPUs or looking to boost speed and scale! This comprehensive guide takes you from novice to expert, showing you how to use Kohya GUI for creating top-notch FLUX LoRAs in the cloud. We'll cover everything: maximizing quality, optimizing speed, and finding the best deals. With our exc...

Zono50•9/11/24, 12:01 PM

@Dr. Furkan Gözükara Flux doesn't support negative prompts does it? i'm making batman but his nipples keep showing for some weird reason

ZZono50 @Dr. Furkan Gözükara Flux doesn't support negative prompts does it? i'm making b...

Furkan Gözükara SECoursesOP•9/11/24, 12:02 PM

it doesnt support by default true. i saw some things at some places but never tried they work or not and dont remembering it

Furkan Gözükara SECoursesOP•9/11/24, 12:02 PM

try to change prompt add more detailed clothing

Furkan Gözükara SECoursesOP•9/11/24, 12:02 PM

armor

Zono50•9/11/24, 12:04 PM

which clip and t5 models are best for text?

Zono50•9/11/24, 12:04 PM

right now i'm using clip-l vit-l-14-text and t5xxl_enconly.safetensors and it's messing up text

ZZono50 right now i'm using clip-l vit-l-14-text and t5xxl_enconly.safetensors and it's ...

Furkan Gözükara SECoursesOP•9/11/24, 12:05 PM

i use fp16 t5 and default clip l for flux

Furkan Gözükara SECoursesOP•9/11/24, 12:06 PM

wget https://huggingface.co/OwlMaster/realgg/resolve/main/flux1-dev.safetensors

wget https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/clip_l.safetensors

wget https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp16.safetensors

wget https://huggingface.co/OwlMaster/realgg/resolve/main/ae.safetensors

wget https://huggingface.co/OwlMaster/realgg/resolve/main/flux1-dev.safetensors

wget https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/clip_l.safetensors

wget https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp16.safetensors

wget https://huggingface.co/OwlMaster/realgg/resolve/main/ae.safetensors

FFurkan Gözükara SECourses dmn model learnt even my broken teeth haah 😄

druhl•9/11/24, 12:29 PM

I found it learnt a mfking mosquito in one photo lol

Ddruhl I found it learnt a mfking mosquito in one photo lol

Furkan Gözükara SECoursesOP•9/11/24, 12:38 PM

magical•9/11/24, 12:52 PM

Hello. Which setting sets the clip l learning rate in the config? Text encoder learning rate?

Mmagical Hello. Which setting sets the clip l learning rate in the config? Text encoder l...

Furkan Gözükara SECoursesOP•9/11/24, 12:57 PM

yes text encoder learning rate

Furkan Gözükara SECoursesOP•9/11/24, 1:22 PM

neo•9/11/24, 1:52 PM

@Dr. Furkan Gözükara DR is there a problem with supir?

neo•9/11/24, 1:53 PM

neo•9/11/24, 2:16 PM

no sorry my bad.. I tried it again on a different instance for the 2nd time and it worked

neo•9/11/24, 2:16 PM

sorry for the pin @

TOPUI•9/11/24, 3:45 PM

Traceback (most recent call last):
File "E:\OneTrainer\OneTrainer\modules\ui\TrainUI.py", line 552, in __training_thread_function
trainer.start()
File "E:\OneTrainer\OneTrainer\modules\trainer\GenericTrainer.py", line 122, in start
self.model = self.model_loader.load(
File "E:\OneTrainer\OneTrainer\modules\modelLoader\FluxLoRAModelLoader.py", line 46, in load
base_model_loader.load(model, model_type, model_names, weight_dtypes)
File "E:\OneTrainer\OneTrainer\modules\modelLoader\flux\FluxModelLoader.py", line 261, in load
raise Exception("could not load model: " + model_names.base_model)
Exception: could not load model: E:/stable-diffusion-webui-forge/webui_forge_cu121_torch231/webui/models/Stable-diffusion/flux1-dev.safetensors

TOPUI•9/11/24, 3:47 PM

I have this model, but when I start training it says there is no model @Dr. Furkan Gözükara

PixelMonkey•9/11/24, 4:52 PM

Is there any nice Lora for fixing hands and faces in Flux you guys can recommend? Especially when the character is further away form the cam i get issues .

PixelMonkey•9/11/24, 4:53 PM

FFurkan Gözükara SECourses Click to see attachment

xlads•9/11/24, 5:45 PM

this looks amazing

FFurkan Gözükara SECourses Click to see attachment

xlads•9/11/24, 5:48 PM

yes I've noticed also smiling a lot in my generations. Also face tends to be in the middle of frame usually

PPixelMonkey Is there any nice Lora for fixing hands and faces in Flux you guys can recommen...

�

𝔸𝖨ΞӾ•9/11/24, 5:55 PM

I would like to know about this also! One solution that I've found works well and is helpful especially if you know how to draw is to generate a Canny preview and then download and edit the preview in Photoshop. So pretty much erase the lines of the hand and redraw the basic details and then reupload it into Controlnet and that will guide Flux to fill in the rest.

�𝔸𝖨ΞӾ I would like to know about this also! One solution that I've found works well an...

PixelMonkey•9/11/24, 5:57 PM

Thats a nice trick.. I need to do 60+ images for a client.. that might not be in the budget hahha but good trick

. I try atm to use Boreal Lora, and <segment:face,0.7> . But with later i am not sure if it works just like that in SD Forge or i need to add specific add on

AIGambino•9/11/24, 7:09 PM

dr 256 images?!

AIGambino•9/11/24, 7:09 PM

all the same outfit? or various

Nneo sorry for the pin @

Furkan Gözükara SECoursesOP•9/11/24, 8:27 PM

awesome

TTOPUI Traceback (most recent call last): File "E:\OneTrainer\OneTrainer\modules\ui\T...

Furkan Gözükara SECoursesOP•9/11/24, 8:27 PM

onetrainer auto downloads

Furkan Gözükara SECoursesOP•9/11/24, 8:27 PM

you have to use hugging face repo

Xxlads yes I've noticed also smiling a lot in my generations. Also face tends to be in...

Furkan Gözükara SECoursesOP•9/11/24, 8:28 PM

i did another training time to test :d

AAIGambino all the same outfit? or various

Furkan Gözükara SECoursesOP•9/11/24, 8:28 PM

various but not all

Furkan Gözükara SECoursesOP•9/11/24, 8:28 PM

different times as well

nicearoni•9/11/24, 9:18 PM

i thought someone shared some flux style presets but I can find it - ring a bell for anyone?

FFurkan Gözükara SECourses various but not all

sergio_gonzo•9/11/24, 9:41 PM

Dr. Furkan I hope everything is going well, I wanted to ask you if I wanted to train the Arcane style from the Netflix series with Flux, how many images would be necessary? And in the process of doing so, the only thing that would change would be the class, right? I would change to style.

Ssergio_gonzo Dr. Furkan I hope everything is going well, I wanted to ask you if I wanted to t...

Furkan Gözükara SECoursesOP•9/11/24, 9:45 PM

more better

wget https://huggingface.co/OwlMaster/realgg/resolve/main/flux1-dev.safetensors wget https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/clip_l.safetensors wget https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp16.safetensors wget https://huggingface.co/OwlMaster/realgg/resolve/main/ae.safetensors

dmn model learnt even my broken teeth haah 😄

Similar Threads

Similar Threads

Similar Threads