Software Engineering Courses (SECourses)•2y ago

Something strange has happenend to my Adapter Face ID....When I input the number of images it just

Something strange has happenend to my Adapter Face ID....When I input the number of images it just duplicates the picture. I have to choose Batch to get different images. I have seed randomized

DDazzastrous [4090]Something strange has happenend to my Adapter Face ID....When I input the numbe...

Furkan Gözükara SECourses•1/11/24, 7:02 PM

let me test

Furkan Gözükara SECourses•1/11/24, 7:03 PM

number of images generated different. testing batch size

Furkan Gözükara SECourses•1/11/24, 7:03 PM

batch size too

Furkan Gözükara SECourses•1/11/24, 7:03 PM

download latest version

Dazzastrous [4090]OP•1/11/24, 7:04 PM

Iamme•1/12/24, 9:10 AM

Is there any open source platform that will animate the whole body of an image?
So I got the 3d Pixar style image generated with ipadapter and i want the image to sing along with the background music, lip sync doesn't have to be perfect, but I want the whole body to move and not just the head.

IIamme Is there any open source platform that will animate the whole body of an image? ...

Furkan Gözükara SECourses•1/12/24, 2:45 PM

there are 2 second ones

Furkan Gözükara SECourses•1/12/24, 2:46 PM

like stable diffusion video

DDeleted User Hi Guys. As a software developer, I am looking for a UK or US partner who can co...

XavierRecio•1/12/24, 6:05 PM

I am interested

XavierRecio•1/12/24, 6:08 PM

Anyone here familiar with using Flutterflow? or any other nocode app to launch something quick? I would like to find a technical cofounder but I need to launch a MVP fast. I am not experienced enough in software development to launch as quickly as I want to. please reply directly to this post.

AIGambino•1/13/24, 4:59 AM

Doctor, what checkpoint did you use on the most recent linkedin post? It's like a cartoon/GTA style

AAIGambino Doctor, what checkpoint did you use on the most recent linkedin post? It's like ...

Furkan Gözükara SECourses•1/13/24, 3:19 PM

can you give me link which one

AIGambino•1/13/24, 6:55 PM

AIGambino•1/13/24, 6:59 PM

@Furkan Gözükara SECourses https://kaist-viclab.github.io/fmanet-site/

FMA-Net: Flow-Guided Dynamic Filtering and Iterative Feature Refine...

Joint Video Super-Resolution (SR) and Deblurring.

AAIGambino <@205854764540362752> https://kaist-viclab.github.io/fmanet-site/

Furkan Gözükara SECourses•1/13/24, 8:01 PM

wow looking nice

Furkan Gözükara SECourses•1/13/24, 8:01 PM

started watcing

joosheen•1/15/24, 10:20 AM

@Furkan Gözükara SECourses Dr Furkan... i have a problem with pixart. I wanted to use 1024 8 bit (just to try if it will be possible with 6gb vram) and i have such an error: OSError: We couldn't connect to 'https://huggingface.co/' to load this file, couldn't find it in the cached files and it looks like PixArt-alpha/PixArt-XL-2-1024-MS is not the path to a directory containing a file named text_encoder\model-00002-of-00002.safetensors.
Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode. It is after downloading model 00001 of 00002.safetensors. What im doing wrong?

Installation

乂

乂AHM3D乂•1/15/24, 4:08 PM

hi all

乂

乂AHM3D乂•1/15/24, 4:08 PM

anyone have Colab Pro?

乂

乂AHM3D乂•1/15/24, 4:08 PM

I need a help please

Jjoosheen <@205854764540362752> Dr Furkan... i have a problem with pixart. I wanted to use...

Furkan Gözükara SECourses•1/15/24, 11:53 PM

are you using my auto installer? looks like you are giving incorrect model path

乂乂AHM3D乂 anyone have Colab Pro?

Furkan Gözükara SECourses•1/15/24, 11:53 PM

i dont have. using kaggle free

FFurkan Gözükara SECourses are you using my auto installer? looks like you are giving incorrect model path

joosheen•1/16/24, 10:45 AM

yes, your auto-installer :/ I wanted to use 1024 8bit

Jjoosheen yes, your auto-installer :/ I wanted to use 1024 8bit

Furkan Gözükara SECourses•1/16/24, 11:27 AM

ok then it should work. let me test

FFurkan Gözükara SECourses ok then it should work. let me test

joosheen•1/16/24, 12:36 PM

Now i'm installing in folder without space letter. We will see, i will text you if it changed sth

FFurkan Gözükara SECourses i dont have. using kaggle free

乂

乂AHM3D乂•1/16/24, 1:16 PM

yeah, using that now.

FFurkan Gözükara SECourses ok then it should work. let me test

joosheen•1/16/24, 1:40 PM

installing in folder without space letter hasn't changed anything :/

Furkan Gözükara SECourses•1/16/24, 2:37 PM

you are running 1024_8bit.bat right?

Furkan Gözükara SECourses•1/16/24, 2:38 PM

the model path is accuratre

Furkan Gözükara SECourses•1/16/24, 2:38 PM

https://huggingface.co/PixArt-alpha/PixArt-XL-2-1024-MS/tree/main

PixArt-alpha/PixArt-XL-2-1024-MS at main

Furkan Gözükara SECourses•1/16/24, 2:53 PM

ok i think issue is diffusers

Furkan Gözükara SECourses•1/16/24, 2:53 PM

fixing

Furkan Gözükara SECourses•1/16/24, 2:57 PM

testing right now to update patreon post

joosheen•1/16/24, 2:57 PM

Yes, 1024_8bit.bat. Ok, so let me know once you fix it, I can't wait to try PixArt

Furkan Gözükara SECourses•1/16/24, 2:58 PM

but i think your issue is another

Furkan Gözükara SECourses•1/16/24, 2:58 PM

delete your cache folder

Furkan Gözükara SECourses•1/16/24, 2:58 PM

pixart

Furkan Gözükara SECourses•1/16/24, 2:58 PM

C:\Users\King.cache\huggingface\hub\models--PixArt-alpha--PixArt-XL-2-1024-MS

Furkan Gözükara SECourses•1/16/24, 2:58 PM

and wait me to update

Furkan Gözükara SECourses•1/16/24, 3:04 PM

i even deleted cache

Furkan Gözükara SECourses•1/16/24, 3:04 PM

trying again

Jjoosheen Yes, 1024_8bit.bat. Ok, so let me know once you fix it, I can't wait to try PixA...

Furkan Gözükara SECourses•1/16/24, 3:05 PM

Furkan Gözükara SECourses•1/16/24, 3:15 PM

working great right now

Furkan Gözükara SECourses•1/16/24, 3:15 PM

try to reinstall

joosheen•1/16/24, 3:15 PM

ok, i deleted PixArt-XL-2-1024-MS folder from .cache folder. I deleted also from .locks folder in .cache folder. I have some tmp files but dont know is it matters?

joosheen•1/16/24, 3:16 PM

ok, im doing clean install

Jjoosheen ok, im doing clean install

Furkan Gözükara SECourses•1/16/24, 3:19 PM

great let me know

Furkan Gözükara SECourses•1/16/24, 3:19 PM

you have python 3.10 righyt?

joosheen•1/16/24, 3:27 PM

yes, 3.10.11 as u said it's the best right now

joosheen•1/16/24, 10:02 PM

@Furkan Gözükara SECourses Ok, this error after installation 1024_8bit "Some modules are dispatched on the CPU or the disk. Make sure you have enough GPU RAM to fit the quantized model. If you want to dispatch the model on the CPU or the disk while keeping these modules in 32-bit, you need to set load_in_8bit_fp32_cpu_offload=Trueload_in_8bit_fp32_cpu_offload=True and pass a custom device_mapdevice_map to from_pretrainedfrom_pretrained" means that i should use lowvram version (i have 6gb vram)?

Something strange has happenend to my Adapter Face ID....When I input the number of images it just

Similar Threads