Software Engineering Courses (SECourses)•3y ago

wdym? it should just be localhost:7860

PPickledJoi wdym? it should just be localhost:7860

DrippyPablo•5/27/23, 12:01 AM

Can I DM you what I mean?

DrippyPablo•5/27/23, 12:16 AM

Nevermind, I think i'm ok, I just re-ran it and it worked

FFurkan Gözükara SECourses i plan to compare all optimizers in kohya lora

Khang•5/27/23, 4:48 AM

Have you done it yet? I'd love to see what you have found out.

Khang•5/27/23, 4:49 AM

Also, I found your videos on many GitHub pages, such as this one: https://github.com/bmaltais/kohya_ss or Runpod page. I don't think other YouTubers are featured as frequently as you, which demonstrates how high your credibility is.

GitHub

GitHub - bmaltais/kohya_ss

Contribute to bmaltais/kohya_ss development by creating an account on GitHub.

Uundelirium better render and more info

AlmoIppy•5/27/23, 5:15 AM

That doesn't really look like it respected the prompts too well, unless that was some checkpoint that has such specific style and no loras.

AlmoIppy•5/27/23, 5:16 AM

I tried regional prompter and while it worked somewhat with the horizontal/vertical divides it quickly failed when I tried it with masking, especially since I was trying to use loras for specific parts of the regions

AlmoIppy•5/27/23, 5:17 AM

when I tried latent instead of attention mode (which is supposed to separate loras into regions) it rendered something that looked nothing like what I was getting withotu regional prompt or with regional prompt in attention mode

AlmoIppy•5/27/23, 5:17 AM

And yeah, I'd rather avoid using inpaint since the goal is to generate lots of decently looking images straight out of the txt2img with maybe optional pass of batch img2img to refine it or alter the style somewhat.

AAlmoIppy That doesn't really look like it respected the prompts too well, unless that was...

undelirium•5/27/23, 5:21 AM

The problem with my showcase is that my softedge was sloppy. I need to redo my open poses. I sorta patched things together in Photoshop, added characters together one at a time. My softedge is the problem in this instance for clarity.

AlmoIppy•5/27/23, 5:21 AM

so which extension is that?

undelirium•5/27/23, 5:21 AM

I have earlier versions that are much more crisp but they lack my jedi

undelirium•5/27/23, 5:22 AM

Latent couple.

undelirium•5/27/23, 5:22 AM

With it I'm using a custom made mask as shown

AlmoIppy•5/27/23, 5:22 AM

does it need any other extension to work?

undelirium•5/27/23, 5:22 AM

Yes control net

AlmoIppy•5/27/23, 5:22 AM

aside from controlnet

undelirium•5/27/23, 5:22 AM

Nope

AlmoIppy•5/27/23, 5:22 AM

you sure?

undelirium•5/27/23, 5:22 AM

Just those two

undelirium•5/27/23, 5:22 AM

And Lora support o guess which is standard tho in webui

undelirium•5/27/23, 5:23 AM

There might be a Lora extension needed to insert them into prompt with weights. I don't use auot1111 anymore

undelirium•5/27/23, 5:23 AM

I've been on a fork for awhile since I don't trust him to stay active

undelirium•5/27/23, 5:24 AM

I know he's been active AF lately. Competition good thing

undelirium•5/27/23, 5:25 AM

heres an earlier step in the process

00203-1256x768-_exceptional_consistentFactor_v40-DPM_2M_Karras-20-7-highly_detailed_cinematic_rendered.png

undelirium•5/27/23, 5:25 AM

later on, there's some things to figure out as i advanced

00237-1256x768-_favorites_revAnimated_v121-DPM_2S_a_Karras-30-10.5-highly_detailed_cinematic_rendered.png

undelirium•5/27/23, 5:25 AM

00239-1256x768-_exceptional_theallysMixIV_v10-DPM_2S_a_Karras-30-10.5-highly_detailed_cinematic_rendered.png

undelirium•5/27/23, 5:25 AM

it does very well, these were made off open poses + softedge nets

undelirium•5/27/23, 5:26 AM

i ended up ditching open pose when i made some spacing changes and added another character, now i need to go back and manually readd the open poses

AlmoIppy•5/27/23, 5:26 AM

That does look better

undelirium•5/27/23, 5:26 AM

none of the detect open poses work very well on mulitple subnjects

AlmoIppy•5/27/23, 5:26 AM

I was thinking of antoher extension that apprently need somethign else to work

AlmoIppy•5/27/23, 5:26 AM

https://github.com/opparco/stable-diffusion-webui-composable-lora

GitHub

GitHub - opparco/stable-diffusion-webui-composable-lora: This exten...

This extension replaces the built-in LoRA forward procedure. - GitHub - opparco/stable-diffusion-webui-composable-lora: This extension replaces the built-in LoRA forward procedure.

undelirium•5/27/23, 5:27 AM

yes that

undelirium•5/27/23, 5:27 AM

that i think is standard in the fork i use i think? idr

undelirium•5/27/23, 5:27 AM

but it lets you do prompts <lora:0.5>

undelirium•5/27/23, 5:27 AM

style weighting

AlmoIppy•5/27/23, 5:27 AM

apparently it says it's compatible with Composable-Diffusion but I can't find what that is

AlmoIppy•5/27/23, 5:27 AM

and I swear I read somewhere that you need both for it to work

undelirium•5/27/23, 5:28 AM

composable is already part of it all im pretty sure, its just standard, the bones of webui

AlmoIppy•5/27/23, 5:28 AM

if you don't use a1111 what do you use?

AlmoIppy•5/27/23, 5:28 AM

whatever is less buggy I'll take it

AlmoIppy•5/27/23, 5:28 AM

spent last week trying to get it to work

undelirium•5/27/23, 5:28 AM

its auto1111 derived so its more or less the same, def less buggy imo but no guarantees

undelirium•5/27/23, 5:28 AM

i use https://github.com/vladmandic/automatic

GitHub

GitHub - vladmandic/automatic: Opinionated fork/implementation of S...

Opinionated fork/implementation of Stable Diffusion - GitHub - vladmandic/automatic: Opinionated fork/implementation of Stable Diffusion

undelirium•5/27/23, 5:29 AM

and it comes with themes baked in so you dont have to use the orange on black

undelirium•5/27/23, 5:29 AM

https://github.com/anapnoe/stable-diffusion-webui-ux

GitHub

GitHub - anapnoe/stable-diffusion-webui-ux: Stable Diffusion web UI UX

Stable Diffusion web UI UX. Contribute to anapnoe/stable-diffusion-webui-ux development by creating an account on GitHub.

undelirium•5/27/23, 5:29 AM

this one is another good alternatiuve too

undelirium•5/27/23, 5:29 AM

the maintainers love their work and practice better coding best practices than auto

undelirium•5/27/23, 5:30 AM

and compatible withall extensions

AlmoIppy•5/27/23, 5:30 AM

and now lycoris loras dont' work sometimes, just gives out an error: RuntimeError: The size of tensor a (9) must match the size of tensor b (4) at non-singleton dimension 1

wdym? it should just be localhost:7860

Similar Threads