Software Engineering Courses (SECourses)•3y ago

Looks like a Network Volume error, you need to increase your Network Volume size if that happens.

xlads•7/14/23, 7:06 PM

im also getting this error when i try to click on an image File Load Error for sample_980-1.png
Unhandled error

xlads•7/14/23, 7:06 PM

inside of juypter on runpod

Aashley Looks like a Network Volume error, you need to increase your Network Volume size...

xlads•7/14/23, 7:09 PM

how can i do this

Xxlads how can i do this

ashleyOP•7/14/23, 7:11 PM

GPU Cloud -> My Storage -> Find Network Volume -> Click Pencil icon to Edit it -> Increase size -> Save

xlads•7/14/23, 7:12 PM

I have nothing in my storge tab i do in "MYPOD" tho

Xxlads I have nothing in my storge tab i do in "MYPOD" tho

ashleyOP•7/14/23, 7:15 PM

Then you don't have a network volume, can you DM me a screenshot of your pod

xlads•7/14/23, 7:35 PM

it doesnt charge more when i increse the disk?

Xxlads it doesnt charge more when i increse the disk?

ashleyOP•7/14/23, 7:40 PM

It does, you pay per GB

ashleyOP•7/14/23, 7:42 PM

https://docs.runpod.io/docs/faq#what-are-pods

RunPod

FAQ

General questions about RunPod and its services.

xlads•7/14/23, 7:53 PM

is it normal?

xlads•7/14/23, 7:54 PM

im just trying to run dreambooth training

Xxlads im just trying to run dreambooth training

ashleyOP•7/14/23, 8:05 PM

What did you set your save model frequency to?

xlads•7/14/23, 8:07 PM

what is model frequency ?

Furkan Gözükara SECourses•7/14/23, 8:30 PM

time to record comfyui tutorial

Deleted User•7/14/23, 8:57 PM

anyone got sdxl working on vast rather than runpod?

Xxlads have you done this?

DeadMan1999•7/14/23, 9:23 PM

no, i just shared the news.

theUnlikely•7/14/23, 9:52 PM

Has anyone had luck using the --sd_model parameter in "Prompts from file or textbox"? According to the Automatic1111 wiki, it's supported, but I get an error when trying to use it.

KKhang @Dr. Furkan Gözükara I found this very interesting extension Dr.Furkan https://g...

Furkan Gözükara SECourses•7/14/23, 10:36 PM

nice

DDeadMan1999 https://huggingface.co/papers/2307.06949 looks at this guys!

Furkan Gözükara SECourses•7/14/23, 10:36 PM

yep i saw this but no code

DDeleted User anyone got sdxl working on vast rather than runpod?

Furkan Gözükara SECourses•7/14/23, 10:37 PM

i made a video about how to use on colab

Furkan Gözükara SECourses•7/14/23, 10:37 PM

just finished recording

Dragonasu•7/14/23, 10:39 PM

Hi guys

Dragonasu•7/14/23, 10:39 PM

Anyone have newsletter?

DDragonasu Anyone have newsletter?

Furkan Gözükara SECourses•7/14/23, 10:40 PM

i have my own on linked in :d

Furkan Gözükara SECourses•7/14/23, 10:41 PM

https://www.linkedin.com/newsletters/technology-and-programming-7018591715401842688/

Technology and Programming | LinkedIn

Furkan Gözükara | Stay on top of the latest tech trends and programming practices with our insightful newsletter. #technology #programming

Furkan Gözükara SECourses•7/14/23, 10:52 PM

testing all loras haha 192 queue

Furkan Gözükara SECourses•7/14/23, 10:52 PM

FFurkan Gözükara SECourses Click to see attachment

ashleyOP•7/14/23, 10:59 PM

Looks like you are actually enjoying Comfy UI now

graphsaregood8626•7/14/23, 11:20 PM

Has anyone tried hosting their own SD models through their own domain?

I'm trying to build something a bit more consumer-friendly, so less gradio-like and maybe more stability AI like. Or maybe a1111+Gradio should be the starting point?

Any pointers are much appreciated! I'm super weak on web dev

Thanks!

Ggraphsaregood8626 Has anyone tried hosting their own SD models through their **own** domain? I'...

ashleyOP•7/14/23, 11:23 PM

Why would you want to do this? You would need multiple GPU for many different people to be able to access it simultaneously.

Aashley Why would you want to do this? You would need multiple GPU for many different pe...

graphsaregood8626•7/14/23, 11:25 PM

Yes, hoping to eventually build out consumer product. Not just giving away free resources to everyone

Ggraphsaregood8626 Yes, hoping to eventually build out consumer product. Not just giving away free...

ashleyOP•7/14/23, 11:30 PM

It is not a trivial task, definitely not something I would want to do.

FFurkan Gözükara SECourses i made a video about how to use on colab

Deleted User•7/14/23, 11:41 PM

I dont have any compute units on colab, was wondering if it would be possible on vast instead

FFurkan Gözükara SECourses no idea atm 😄

monsieurBoutte•7/15/23, 12:31 AM

https://youtu.be/ot5GkaxHPzk

Ideally I could stand up something like this behind an API

YouTubeSECourses

Stable Diffusion Now Has The Photoshop Generative Fill Feature With...

Adobe changed the game forever with Firefly Photoshop Generative Fill feature. However, Adobe subscription is extremely expensive. As of today, Stable Diffusion Automatic1111 Web UI ControlNet extension has the same generative fill feature. In this video, I will show you how you can use Mikubill ControlNet generative fill like Photoshop has for ...

MmonsieurBoutte https://youtu.be/ot5GkaxHPzk Ideally I could stand up something like this behin...

Furkan Gözükara SECourses•7/15/23, 12:32 AM

auto1111 has roop extension

Furkan Gözükara SECourses•7/15/23, 12:32 AM

and api support

EyeSpyBekoAI•7/15/23, 1:18 AM

Hey there SD community. I had a hopefully somewhat quick question for you I'm guessing one of you can answer without an issue!
I was wondering if you guys know a way to put a generated image from SD, infront of a pure white background. An example is below. I am able to do this right now by first uploading a pure white background into inpaint, mask an area, and then generate an image. Which works well enough, but if I want to use the shape of a specific image and get control net involved, the quality takes a big hit, and then the pure white background disappears. Do any of you know a way I can do this in a much more straightforward fashion? Any help would be great. An example is below. Thanks!

weimaraner-dog-pure-white-background-cute-young-33321196.png

EyeSpyBekoAI•7/15/23, 1:23 AM

Oh and if it helps, I have the rembg extension installed and can remove the background of a generated picture if needed. It’s then a question of how to replace it with a pure white background.

EEyeSpyBekoAI Oh and if it helps, I have the rembg extension installed and can remove the back...

Furkan Gözükara SECourses•7/15/23, 1:29 AM

as far as i know there is no way

Furkan Gözükara SECourses•7/15/23, 1:30 AM

except if you train object

Furkan Gözükara SECourses•7/15/23, 1:30 AM

you are already doing background remove

FFurkan Gözükara SECourses as far as i know there is no way

EyeSpyBekoAI•7/15/23, 1:31 AM

So train the objects with pure white backgrounds?
I tried training on no backgrounds, and it had really bad results SD got really confused.

EEyeSpyBekoAI So train the objects with pure white backgrounds? I tried training on no backgro...

Furkan Gözükara SECourses•7/15/23, 1:35 AM

stable diffusion dont have transparent layer

Furkan Gözükara SECourses•7/15/23, 1:35 AM

it is parsed as white or black

Furkan Gözükara SECourses•7/15/23, 1:36 AM

https://www.linkedin.com/posts/furkangozukara_testing-different-sdxl-lora-trainings-different-activity-7085794214491435008-2iTN?utm_source=share&utm_medium=member_desktop

Furkan Gözükara on LinkedIn: Testing different SDXL LoRA trainings ...

Testing different SDXL LoRA trainings (different configurations & checkpoints) for hopefully upcoming amazing tutorial

Full workflow will be hopefully…