Software Engineering Courses (SECourses)•3y ago

Okay for 5hrs working on this video following this settings for 8bitAdaw with exact parameters. Well

Okay for 5hrs working on this video following this settings for 8bitAdaw with exact parameters. Well, it did not get me precise results on my subject images "face" it looks like the subject but not good at all. Therefore settings from other may not work for others. Im thinking it dependes on what gpu we are running. Im running rtx 3060 12gb latest drivers updated. The man used higher graphics.

Vvjavad1443 @Dr. Furkan Gözükara Hi Dr, I'm truly appreciate for your time and effort. I've ...

Furkan Gözükara SECourses•3/8/23, 10:35 AM

i say yes use xformers 0.0.17 and ema

Bbush i have a question. Anyone using dreambooth must've come across JOE PENNA dreambo...

Furkan Gözükara SECourses•3/8/23, 10:36 AM

i havent checked. but you can use only man or photo of man as classification. it works even without classification but i found classification improves styling capability for faces

Vvjavad1443 Does anybody know why I'm stuck in the loop when run "python relauncher.py" comm...

Furkan Gözükara SECourses•3/8/23, 10:37 AM

change the loop in relauncher.py

TTrimac Okay for 5hrs working on this video following this settings for 8bitAdaw with ex...

Furkan Gözükara SECourses•3/8/23, 10:38 AM

could be related to your training dataset

Furkan Gözükara SECourses•3/8/23, 10:38 AM

no backgrounds clothes should be same

TTrimac Okay for 5hrs working on this video following this settings for 8bitAdaw with ex...

Furkan Gözükara SECourses•3/8/23, 10:47 AM

my computer also has 12 gb rtx 3060. it really looks good. you should be able to get good results same as base test on that video

FFurkan Gözükara SECourses i say yes use xformers 0.0.17 and ema

vjavad1443•3/8/23, 10:51 AM

So in youtube video you said the address we should run those commands ( uninstall ) is located in /workspace/venv/lib/python3.10

but for me there is no venv in workspace folder

I can find it in this address:
/workspace/stable-diffusion-webui/venv/lib/python3.10

vjavad1443•3/8/23, 10:51 AM

Is that ok?

Vvjavad1443 So in youtube video you said the address we should run those commands ( uninstal...

Furkan Gözükara SECourses•3/8/23, 10:51 AM

in unix

Furkan Gözükara SECourses•3/8/23, 10:51 AM

it works wherever you run actually

Furkan Gözükara SECourses•3/8/23, 10:51 AM

i mean on runpod at least

Furkan Gözükara SECourses•3/8/23, 10:51 AM

so it is okay

Furkan Gözükara SECourses•3/8/23, 10:52 AM

i think it is because of the template

Furkan Gözükara SECourses•3/8/23, 10:52 AM

template links to venv so all command run on venv automatically

vjavad1443•3/8/23, 10:55 AM

yes, I'm working on runpod. also I see some changes. For example I selected SD v1.5 template & after I opened it I see there is two models :

SDv1-5.ckpt
SDv2-768.ckpt

Furkan Gözükara SECourses•3/8/23, 10:56 AM

they changed template

Furkan Gözükara SECourses•3/8/23, 10:56 AM

but still works very well

Furkan Gözükara SECourses•3/8/23, 10:56 AM

currently i am doing 3 experiment

Furkan Gözükara SECourses•3/8/23, 10:56 AM

on dreambooth

Furkan Gözükara SECourses•3/8/23, 10:56 AM

on that template

vjavad1443•3/8/23, 10:57 AM

and would you please answer to this question as well. your opinion is important to me

I wanted to know have you tried the below format for learning rate field instead of 0.00001 ?

0.05:10,0.02:20,0.01:60,0.005:200,0.002:500,0.001:3000,0.0005

Vvjavad1443 and would you please answer to this question as well. your opinion is important ...

Furkan Gözükara SECourses•3/8/23, 11:02 AM

i am using constant

Furkan Gözükara SECourses•3/8/23, 11:02 AM

that can train very quickly

vjavad1443•3/8/23, 11:06 AM

I got this error

Screen_Shot_2023-03-08_at_4.05.17_AM.png

vjavad1443•3/8/23, 11:07 AM

I run command the first for command without any issue

Screen_Shot_2023-03-08_at_4.06.42_AM.png

vjavad1443•3/8/23, 11:07 AM

I run the first four command without any issue *

FFurkan Gözükara SECourses i havent checked. but you can use only man or photo of man as classification. i...

bush•3/8/23, 11:08 AM

i was actually talking about the content of the images and not the name of class

FFurkan Gözükara SECourses but this machine consumes a lot credits

Arvy•3/8/23, 11:09 AM

I had my 40gb vram card run out of memory several times yesterday on Runpod. I switched to vast.ai and haven't faced an issue so far. However downloading from vast.ai is very slow.

bush•3/8/23, 11:10 AM

like here, they arwe all real photos from unsplash

vjavad1443•3/8/23, 11:11 AM

This command didn't work as well

Screen_Shot_2023-03-08_at_4.11.13_AM.png

Bbush i was actually talking about the content of the images and not the name of class

Furkan Gözükara SECourses•3/8/23, 11:15 AM

the images are generated by the name of class

AArvy I had my 40gb vram card run out of memory several times yesterday on Runpod. I s...

Furkan Gözükara SECourses•3/8/23, 11:15 AM

it totally depends on configuration

Bbush like here, they arwe all real photos from unsplash

Furkan Gözükara SECourses•3/8/23, 11:16 AM

if you want to fine tune model

Furkan Gözükara SECourses•3/8/23, 11:16 AM

improve quality

Furkan Gözükara SECourses•3/8/23, 11:16 AM

then you can feed better images

Furkan Gözükara SECourses•3/8/23, 11:16 AM

the classification image logic is

Furkan Gözükara SECourses•3/8/23, 11:16 AM

harder to explain quickly

Furkan Gözükara SECourses•3/8/23, 11:16 AM

but i am planning a video for that

Vvjavad1443 This command didn't work as well

vjavad1443•3/8/23, 11:17 AM

I used this command instead and it worked:

pip install xformers==0.0.17.dev464

Is it ok ?

Bbush like here, they arwe all real photos from unsplash

bush•3/8/23, 11:17 AM

gotcha. also, one more thing. so I have the deleted the pod on which i trained yesterday. but now I want to fine tune the model even more. the steps will be the same?

Vvjavad1443 I used this command instead and it worked: pip install xformers==0.0.17.dev464 ...

Furkan Gözükara SECourses•3/8/23, 11:18 AM

yep

Bbush gotcha. also, one more thing. so I have the deleted the pod on which i trained y...

Furkan Gözükara SECourses•3/8/23, 11:19 AM

you need to upload same model

Furkan Gözükara SECourses•3/8/23, 11:19 AM

to continue training

Furkan Gözükara SECourses•3/8/23, 11:19 AM

or it will beging from 0

bush•3/8/23, 11:20 AM

oh yes that i'll upload, do i select this model as the base model in dreamboot and then create another new dreambooth model?