Software Engineering Courses (SECourses)•3y ago

anyone had massive it/s slowdown on the new 1.3.0 a1111 commit? reverting commits didn't give me bac

anyone had massive it/s slowdown on the new 1.3.0 a1111 commit? reverting commits didn't give me back my old speed either and i'm trying to figure out what might've caused this

FFurkan Gözükara SECourses but on unix i couldnt yet

ashley•5/29/23, 5:32 AM

I also couldn't get the latest version of torch working on Linux

Tthebean anyone had massive it/s slowdown on the new 1.3.0 a1111 commit? reverting commit...

ashley•5/29/23, 5:36 AM

It shouldn't make any difference to things like that. It's usually your GPU that affects things like that.

krispydonut•5/29/23, 5:36 AM

Hi, need your guys help i tried using stable diffusion on runpod and have downloaded the control net extension but somehow its not working. i use the control net extension and have downloaded the relevant files

@Furkan Gözükara SECourses

Aashley It shouldn't make any difference to things like that. It's usually your GPU that...

thebeanOP•5/29/23, 5:50 AM

well there's quite a few software versions and settings that can massively affect speed in a1111. i didn't change any of that besides doing a git pull though, and i didn't do any operating system updates on my linux system which is why i'm confused

thebeanOP•5/29/23, 5:50 AM

downgrading to my old webui commit didn't help either ; /

生

生死•5/29/23, 7:57 AM

when i run the code in this vedio .cmd show this to me. is that means i dont run on the GPU?

生

生死•5/29/23, 7:57 AM

https://www.youtube.com/watch?v=ku6UvK1bsp4&t=631s

YouTubeSECourses

Dolly 2.0 : Free ChatGPT-like Model for Commercial Use - How To Ins...

Databricks’ #Dolly v2 is a free, open source, commercially useable ChatGPT-style #AI model. Dolly 2.0 could spark a new wave of fully open source LLMs similar to #ChatGPT . Open source community working hardest to bring up a model that can compete with GPT4. Our discord: https://bit.ly/SECoursesDiscord

If I have been of assistance to you and yo...

生

生死•5/29/23, 7:57 AM

生

生死•5/29/23, 7:58 AM

i look my Task Manager my ram is increasing so high but gpu not increase

Fabricatedgirls•5/29/23, 8:20 AM

do you need regularization images for training a photograph style lora from movie references?
or is there a good tutorial specifically for this?
its unclear when you need to tag the images, (or remove stuff from tags) when you need regularization images, when it comes to training a photographic style vs a character vs a pose/action vs details applied to any of the above, is there good info about this anywhere? (info seems to change every 2 weeks lol)
or a loha/lycoris if they are better
never tried to train one

Fabricatedgirls•5/29/23, 8:24 AM

like a table that says:
training photoreal faces? dont need regularisation images, here's good starter settings
training a photographic style, dont do this, these settings are good, etc, would be amazing

Fabricatedgirls•5/29/23, 8:25 AM

that could be kept updated

Θεόδωρος🔱•5/29/23, 9:48 AM

are these good settings for dreambooth

Θεόδωρος🔱•5/29/23, 9:49 AM

also for training people is it best to use LoRa or Dreambooth training?

Θεόδωρος🔱•5/29/23, 9:49 AM

so for the best resultst possible

Θεόδωρος🔱•5/29/23, 9:49 AM

size doesnt matter for me

ΘΘεόδωρος🔱also for training people is it best to use LoRa or Dreambooth training?

ashley•5/29/23, 10:04 AM

Dr. Furkan Gözükara told me that Dreambooth is the best for training people. I haven't actually tried LoRA yet, but Textual Inversion produced terrible results. Dreambooth produces great results.

Θεόδωρος🔱•5/29/23, 10:10 AM

Ok so good choice i chose dreambooth

Θεόδωρος🔱•5/29/23, 10:10 AM

Thank you very much for your repsonse

Tthebean anyone had massive it/s slowdown on the new 1.3.0 a1111 commit? reverting commit...