Software Engineering Courses (SECourses)•16mo ago

started new training :d

FFurkan Gözükara SECourses started new training :d

Vinmi•9/12/24, 1:31 AM

Vinmi•9/12/24, 1:32 AM

also training here

Vinmi•9/12/24, 1:32 AM

but on the cloud, my 3070TI has a shitty amount of VRAM for that task

VVinmi Click to see attachment

Furkan Gözükara SECoursesOP•9/12/24, 1:34 AM

use massed compute

Furkan Gözükara SECoursesOP•9/12/24, 1:34 AM

way cheaper

Furkan Gözükara SECoursesOP•9/12/24, 1:34 AM

way better quality with our config

Vinmi•9/12/24, 1:34 AM

I had some credits there

Furkan Gözükara SECoursesOP•9/12/24, 1:34 AM

h100 total waste of money :d

Furkan Gözükara SECoursesOP•9/12/24, 1:35 AM

for example lora rank 16 = much lower quality compared to 128

Furkan Gözükara SECoursesOP•9/12/24, 1:35 AM

i compared

FFurkan Gözükara SECourses for example lora rank 16 = much lower quality compared to 128

Vinmi•9/12/24, 1:53 AM

I'm training again on 128

AIGambino•9/12/24, 3:00 AM

what quality specifically dr?

Vinmi•9/12/24, 3:37 AM

did it and now all I get are big headed photos

Dragoy•9/12/24, 3:57 AM

@Lecturer - Main Admin I wonder if you've ever used modal.com? How difficult is it to run kohya_ss there?

xlads•9/12/24, 5:16 AM

xlads•9/12/24, 5:19 AM

Something that I see has helped in prompts that include multiple people is changing what the people are wearing their age and hair color

mikemenders•9/12/24, 5:25 AM

I'm still waiting with Flux for several reasons. For one thing, what I see from current loras is that low weight/alpha is enough for posing, so some people use 8s. On the other hand, human anatomy is horrible, so not trained on nsfw images, so it will be a lot more work to train all body parts, and I'd rather wait for that until more sophisticated flux models are available. As I see it, it will be at least 2-3 months. The Pony is also being made for Auraflow first (which is also because of the license), so the transition to flux is a long way off for me.

wardensc2•9/12/24, 6:37 AM

Does fp8 vs fp16 lora quality diffrerent much ?

DDragoy @Lecturer - Main Admin I wonder if you've ever used modal.com? How difficult is ...

Furkan Gözükara SECoursesOP•9/12/24, 9:36 AM

no idea

AAIGambino what quality specifically dr?

Furkan Gözükara SECoursesOP•9/12/24, 9:36 AM

quality of training

Wwardensc2 Does fp8 vs fp16 lora quality diffrerent much ?

Furkan Gözükara SECoursesOP•9/12/24, 9:37 AM

a little bit

magical•9/12/24, 10:09 AM

For sdxl training on runpod, restarting the pod or pkill -f "relauncher.py" isn't ending the gpu memory usage for me

magical•9/12/24, 10:09 AM

Does anyone else have this problem?

magical•9/12/24, 10:20 AM

but if I only want to train and not use A1111 in kohya I don't have to choose Stable Diffusion template and upload the relauncher.py, right?

TOPUI•9/12/24, 10:54 AM

If my current image size is 16:9 and the resolution is 1024X576, is this how the size should be filled in on OneTrainer? @Dr. Furkan Gözükara

hellolaco•9/12/24, 10:56 AM

if I want to run Flux dev on a cloud with Forge webui, which site do you recommend? I've tried runpod but their Forge is not updated.

liorftw•9/12/24, 11:45 AM

https://github.com/rickrender/FluxLatentUpscaler
hey i saw this, and i trained a lora model with my dataset, can i use this upscaler in the swarmui? because the results looks very good, or it's something not relatable and its his own model

GitHub

GitHub - rickrender/FluxLatentUpscaler: This Flux latent upscaler w...

This Flux latent upscaler workflow creates a lower-resolution initial pass, then advances to a second pass that upscales in latent space to twice the original size. Latent space manipulations in th...

Mmagical For sdxl training on runpod, restarting the pod or pkill -f "relauncher.py" isn'...

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

I overwrite it and then do fuser -k 3000/tcp

Mmagical but if I only want to train and not use A1111 in kohya I don't have to choose St...

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

Use our kohya installer on torch 2.1 template

TTOPUI If my current image size is 16:9 and the resolution is 1024X576, is this how the...

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

I would ask this to them I don't know how exactly they handle

Hhellolaco if I want to run Flux dev on a cloud with Forge webui, which site do you recomme...

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

Massed compute

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

Image updated work out of box

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

Select our image SECourses use our coupon SECourses for 31 cents per hour on a6000

Lliorftw https://github.com/rickrender/FluxLatentUpscaler hey i saw this, and i trained a...

Furkan Gözükara SECoursesOP•9/12/24, 12:13 PM

Nothing is as good as supir have you tested it?

Furkan Gözükara SECoursesOP•9/12/24, 12:13 PM

But it is seperate app sadly not in swarmui

Furkan Gözükara SECoursesOP•9/12/24, 12:13 PM

My strategy is select images and upscale as batch

FFurkan Gözükara SECourses I overwrite it and then do fuser -k 3000/tcp

sergio_gonzo•9/12/24, 12:56 PM

Dr Furkan i hope everything is ok, im wondering now that you have been making test with OneTrainer who do you thinbk could be more fast to make model in local between kohya and OneTrainer?

wardensc2•9/12/24, 1:13 PM

dear @Dr. Furkan Gözükara what does the "segment prompt" meaning. I see in your test prompt:
photo of ohwx man a red shiny steel armor with scale patterns like a dragon scales in an epic battlefield as the leader of the army and holding a mighty gigantic diamond sword , face open looking at the camera on a bright sunny day, epic scene, cinematic shot<segment:face,0.7>photo of ohwx man

wardensc2•9/12/24, 1:14 PM

is that we ask Comfy or Forge to do the face detailer without setting node or run addetailer extension