Skip to content

adriangalilea/LambdaTunnel

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

50 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Why?

Performance for stable-diffusion on MacOS/iPadOS is not that great, many people are building dedicated Windows systems with high-end GPU's, and that's great but I think we can do better.

What about using locally a remote Nvidia H100 ($40,000 GPU) at $1,90 per hour?

Features

image
See what is available and directly launch an instance.

image
Create an alarm for a specific type of instance.

image
Launch an instance.

image
Use the instance.

image
You can launch a payload that will run everything for you, including the ssh tunnel. I will add more than Fooocus, but currently is the one I'm using, feel free to add others.
It'll automatically open a tab on your browser when it's done installing everything and doing the tunnel.

image

image

Prerequisites

  • An account on Lambda Cloud.

Steps

  1. CLone the repo
  2. install requirements.txt
  3. Run main.py

Lamdalabs

I'm not affiliated in any way shape or form to Lambdalabs.

WARNING $$$

BE SURE TO CLOSE THE INSTANCE WHEN YOU ARE DONE I WOULD LIKE TO BE ABLE TO CLOSE THE SESSION BY SHUTTING DOWN THE SERVER BUT LAMBDALABS DON'T ALLOW THIS, I'M SHOUTING IN HOPES YOU SAVE SOME HARD EARNED CASH, IF YOU ARE READING THIS AFTER HAVING SPENT A BUNCH OF MONEY BECAUSE YOU WENT TO SLEEP AND FORGOT, I'M SORRY BUT THAT'S YOUR FAULT <3, THANKS FOR LISTENING TO MY TED-TALK.

Notice you'll pay for the setup time too, this is meant for serious work or enthusiasts.

Security disclaimer and agreement

Running scripts directly from the web is a potential security risk. Before running the command, you should inspect the script from this repository. By running the command, you are acknowledging that you trust the source and the contents of the script.

While every effort has been made to ensure the safety and effectiveness of this script, it is provided as-is without any warranties or assurances of any kind. The author of this script cannot be held responsible for any damages, data loss, or issues of any kind arising from the use of this script.

To-do

  • Working fast_setup.sh
  • update to python interactive program
  • create a payload for automatic1111
  • upgrade into a textual GUI
  • #1
  • Having an auto-close instance timer so you don't burn money if you forget.
  • YouTube video tutorial
  • Illustrate this readme with performance metrics.

Contributing

I welcome contributions to this project! Please feel free to open an issue or submit a pull request.

Author: Adrian Galilea Co-author: ChatGPT & copilot 🫀x🤖

About

Run Stable Diffusion Automatic1111 webui on a rented H100 but use it locally from Mac or iPad

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages