Software Engineering Courses (SECourses)•3y ago

22:03 How to upload generated checkpoints / model files into Hugging Face for blazing fast upload an

22:03 How to upload generated checkpoints / model files into Hugging Face for blazing fast upload and download
25:02 Where to find Hugging Face uploaded models after upload has been completed

Mmtnmecca train using the methods described and in video by @Dr. Furkan Gözükara but add a...

DR.Siva•12/6/23, 4:34 PM

regularization pics ?

FFurkan Gözükara SECourses 22:03 How to upload generated checkpoints / model files into Hugging Face for bl...

GravNolan•12/6/23, 4:36 PM

i am using run pod, network volume, will it work for that too?

GGravNolan i am using run pod, network volume, will it work for that too?

Furkan Gözükara SECoursesOP•12/6/23, 4:37 PM

it works everywhere

Furkan Gözükara SECoursesOP•12/6/23, 4:37 PM

google colab, your windows pc, kaggle doesnt matter

Furkan Gözükara SECoursesOP•12/6/23, 4:38 PM

on runpod so easy open a juypter tab just like new terminal

DDR.Siva regularization pics ?

mtnmecca•12/6/23, 4:41 PM

i haven't tried two different regularization sets yet. I have trained either man or woman or style and never two different ones at a time. You can combine the trained ckpts and or loras after you train each individually.

DR.Siva•12/6/23, 4:46 PM

have like 20 hr left on kaggle. in your advice , i would use two folders with no reg pics

DR.Siva•12/6/23, 4:47 PM

it will produice only one lora right , i mean i lora set(16 epochs) of two faces..

DR.Siva•12/6/23, 4:48 PM

or else is it possible to generate couple pics and use them for regularization ?

DR.Siva•12/6/23, 4:48 PM

if so , how will the ai know which is male and which is female ? and how would it equate it to our malke and female pics ?

thewingag•12/6/23, 4:50 PM

hey who u can insta magic animate? can u share me video? thanks

thewingag•12/6/23, 4:50 PM

J.•12/6/23, 5:11 PM

@Dr. Furkan Gözükara Any videos of yours to recommend?

Im getting into the whole sdxl / stable diffusion stuff, through fooocus, and learning abotu LORAs, But seems like there is a whole lot more from dreambooth, textual params, control net, (so many ways to modify the model)

DDR.Siva it will produice only one lora right , i mean i lora set(16 epochs) of two faces...

mtnmecca•12/6/23, 5:13 PM

I would train ohwx man in one session with man regularization images and then train again with bbuk woman with woman regularization images. When you are done you can combine the ckpts and or extract loras for each and then use the prompt I listed above. You do not need to have the man and woman together to begin with as a couple.

FFurkan Gözükara SECourses it works everywhere

GravNolan•12/6/23, 5:13 PM

sorry, for disturbing you, but i tried using code from video in terminal it didn't worked.

GravNolan•12/6/23, 5:14 PM

i have set up hugging face and token but the code seems not working.

GravNolan•12/6/23, 5:16 PM

GravNolan•12/6/23, 5:16 PM

this code i have tried.

Mmtnmecca I would train ohwx man in one session with man regularization images and then tr...

DR.Siva•12/6/23, 5:19 PM

since i have trained both of our faces , i have now 16x2 loras. In this i have selected only like 3 loras for each face which are good. IN this i will select the best , next what ?

DDR.Siva since i have trained both of our faces , i have now 16x2 loras. In this i have s...

mtnmecca•12/6/23, 5:24 PM

if you have trained loras already try the procedure I outlined above with adetailer
it takes a lot of trial and error to produce the results you want. Generate lots of images and then refine on the ones you like best

DR.Siva•12/6/23, 5:25 PM

ok will try it.. thanks a lot .I think in the prompt iteself i will tell which side each are .. this might help a bit in each pic

GGravNolan Click to see attachment

DR.Siva•12/6/23, 5:26 PM

check the output folder directory again.

DR.Siva•12/6/23, 5:26 PM

alsowhat is the error you are getting ?

FiveSAI•12/6/23, 5:28 PM

SDXL Lora with new settings, quality quite good

GravNolan•12/6/23, 5:28 PM

it is saying

FFiveSAI SDXL Lora with new settings, quality quite good

GravNolan•12/6/23, 5:28 PM

impressive.

GGravNolan it is saying

DR.Siva•12/6/23, 5:30 PM

you just hit play in that code line irght ?

DR.Siva•12/6/23, 5:30 PM

what was the previoous commands in console ?

FFiveSAI SDXL Lora with new settings, quality quite good

DR.Siva•12/6/23, 5:31 PM

really wish i too had similar results..

DDR.Siva really wish i too had similar results.. 😦

FiveSAI•12/6/23, 5:31 PM

I'll make a tutorial soon, i've done 600+ trainings

FFiveSAI I'll make a tutorial soon, i've done 600+ trainings

GravNolan•12/6/23, 5:32 PM

that will be great.

FiveSAI•12/6/23, 5:32 PM

I found that dreambooth sdxl, has problems with other base models, usually only works with the base sdxl model

FiveSAI•12/6/23, 5:33 PM

even with lora extraction

FFiveSAI even with lora extraction

mtnmecca•12/6/23, 5:45 PM

I've had no issues with realvisxl 2 and juggernautXL (6 or 7)

Mmtnmecca I've had no issues with realvisxl 2 and juggernautXL (6 or 7)

Aldo Raine•12/6/23, 5:47 PM

Is this the correct way to target JuggernautXL v7?
stablediffusionapi/juggernaut-xl-v7

DDR.Siva really wish i too had similar results.. 😦

mtnmecca•12/6/23, 5:50 PM

It might have to do with the training processor specs. I trained on 3060 with 12gb vram and the results were poor to average. I now train with 4090 24 gb vram or higher and the results are substantially better - amazingly actually.

DR.Siva•12/6/23, 5:51 PM

first training i tried was on my 3080 10gb , and when it showed it will take 49 hours , i left and went for runpod , which finished the job around 4-5 hours.. then kaggle has been doing my jobs

Mmtnmecca I've had no issues with realvisxl 2 and juggernautXL (6 or 7)

FiveSAI•12/6/23, 5:56 PM

I mean like dreambooth models i've trained on the base sdxl model, worked super well with sdxl base 1.0, but the results looked weird on real vis, and jug, and others, i mean the quality is good, but the faces sometimes look off

FFiveSAI I mean like dreambooth models i've trained on the base sdxl model, worked super...

mtnmecca•12/6/23, 5:58 PM

interesting - i'll try another training with base 1.0 since its been awhile. thx

Mmtnmecca interesting - i'll try another training with base 1.0 since its been awhile. th...

FiveSAI•12/6/23, 6:01 PM

Might have to do with the text encoder, not sure, but try and let me know

liber vir•12/6/23, 7:01 PM

i am paying 50 dollars if any of u can make me one simple workflow on comfyui with IP adapter, if u are intereseted or if u can help hit me up

virginia•12/6/23, 7:02 PM

i can help you

virginia•12/6/23, 7:02 PM

add me

virginia•12/6/23, 7:02 PM

i am using comfyUI

FFiveSAI I'll make a tutorial soon, i've done 600+ trainings

Benjamin•12/6/23, 7:15 PM

looking forward to have a look into that

Lliber vir i am paying 50 dollars if any of u can make me one simple workflow on comfyui wi...

Benjamin•12/6/23, 7:17 PM

you can find them actually everywhere, e.g. https://github.com/cubiq/ComfyUI_IPAdapter_plus

GitHub

GitHub - cubiq/ComfyUI_IPAdapter_plus

Contribute to cubiq/ComfyUI_IPAdapter_plus development by creating an account on GitHub.

virginia•12/6/23, 7:43 PM

i can only see loras here and i tried, although tags can be added into prompts, it does not work at all. loras should be added from here , really do not know what it is wrong

Dajour•12/6/23, 8:16 PM

Hi everyone. As a full-stack developer, I have rich experiences with image generation platform development with Stable diffusion api. Also chatbot using OpenAI is available. Please DM if someone wants to build awesome platforms. Best Regards.

Tthewingag hey who u can insta magic animate? can u share me video? thanks

Furkan Gözükara SECoursesOP•12/6/23, 8:16 PM

hello. i havent made the video yet

Furkan Gözükara SECoursesOP•12/6/23, 8:17 PM

i can try to help

22:03 How to upload generated checkpoints / model files into Hugging Face for blazing fast upload an

Similar Threads