Furkan Gözükara
patreon
Massed Compute Installers - Upgrade Automatic1111 - Coupon Code - ControlNet - ADetailer - Facefusion - Reactor & More
🕑 Added 2024-05-19 23:26:00 +0000 UTCThis post is obsolete New post : https://www.patreon.com/posts/105735932
This post is obsolete New post : https://www.patreon.com/posts/105735932
-
-
-
Join discord to get help, chat, discuss and also tell me your discord username to get your special rank : SECourses Discord
Register Massed Compute From Below Link (could be necessary to use our Special Coupon for A6000 GPU for 31 cents per hour): https://bit.ly/Furkan-Gözükara
Massed Compute Tutorials
Biggest : Full Stable Diffusion SD & XL Fine Tuning Tutorial With OneTrainer On Windows & Cloud - Zero To Hero
IDM-VTON: The Most Amazing Virtual Clothing Try On Application - RunPod - Massed Compute - Kaggle
Stable Cascade Full Tutorial for Cloud - Predecessor of SD3 - Massed Compute, RunPod & Kaggle
20 May 2024 Update
Files upgraded to V5 : mass_compute_v5.zip
Automatic1111 SD Web UI upgrader / installer file updated
Now when you run Upgrade_Automatic1111 .sh file, it will upgrade Automatic1111 Web UI to the latest version
Moreover, it will automatically install latest version of After Detailer (ADetailer), FaceFusion, Reactor, ControlNet extensions
It will also auto install system wide FFmpeg
To have more information please read the GitHub Readme file
Hugging Face repository downloader code added to the Upload_Hugging_Face .ipynb file
So you can download your private Hugging Face repos 1 click even if they are 100s of GBs
Sadly TensorRT still not working
Our image getting updated. Now includes pre-installed Forge as well as to Kohya, OneTrainer and Jupyterlab
How To Deploy
Please watch above tutorials to learn
Then from here deploy your VM as shown in below image: https://vm.massedcompute.com/deploy
Comments
Furkan Gözükara
RTX 3090 and A6000 different gpus. during inference that could be reason. perhaps try bigger batch size comparison
Furkan Gözükara
Massed compute has A6000 GPU. Not same as RTX 3090 that could be reason.
Josh Tech
Hello! I compared RunPod with 3090 and Massed Compute regarding the Speed in SDXL (Realismengine). I installed your update_sd1111 file and changed the COMMANDLINE_ARGS in the settings. I generated 16 images (4 count x 4 batch) with 1024*1024 with DPM SDE Karras 30 Steps. Runpod took 2.5 minutes and Massed Compute took 3.50 minutes. You can take a look at the settings and the PNG (for reproduction) on my Google Drive. Do you have any idea why massed compute was slower? https://drive.google.com/drive/folders/1KnT4HCRorzAcgDTDgag0SgSJS7kTjbDE?usp=sharing
Furkan Gözükara
sure i will make it hopefully next
Nenad Kuzmanovic
Great tutorial. What i would like (and i believe many others) is to have Comfyui tutorial, to install it on Massed Compute. I know a lot of people suffer from not having good gpu. Either my self, i have 3090 but for more complex animations, i just need A6000.
Furkan Gözükara
i am going to forward this to them
Nenad Kuzmanovic
i don' like the way they are managing payments. On Runpod u can pay with your card and not save card info. Here, they want you to first add card (for future payments) and then you can use it...
Furkan Gözükara
this one has Ubuntu desktop not jupyterlab interface
diffusers
Wow 31 cents?! Thats awesome. Is the setup similar to runpod?
