-
Notifications
You must be signed in to change notification settings - Fork 27.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Hardware Auto-Setup for Examples #22319
Conversation
The documentation is not available anymore as the PR was closed or merged. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks a lot for your PR. While we're super happy to showcase RunHouse examples, I feel having one page in the doc is maybe a step too far at this stage. Could we focus first on:
- having one example as you did
- say a bit more clearly in the README that RunHouse is one way to run the examples remotely (and there might be others)
- make launch_auto_hardware.mdx a notebook you host on your side and link to it from the examples/README.md
What so you think?
Definitely, that makes sense, and thanks for the fast review! Updated per your suggestions. |
Thanks for iterating! There is just the issue of the tests that are not running now. It seems there is an issue with your CircleCI permissions. Could you try refreshing your permissions as shown here? |
Oh sorry! I think I've granted it access. Do I need to trigger anything on my side? |
Probably an empty commit to re-trigger the CI. |
Ok, so it looks like we just need a quick |
Oops, thought I already ran via |
Ok, last failures are fixed on main so merging. Thanks! |
Thank you, Sylvain!! |
* Add initial remote hardware auto-setup docs * Fix a few typos and clarify some language * Add missing dependency * Update self-hosted launch script with Sylvain's comments. * Formatting. * Trigger CI * Style
* Add initial remote hardware auto-setup docs * Fix a few typos and clarify some language * Add missing dependency * Update self-hosted launch script with Sylvain's comments. * Formatting. * Trigger CI * Style
What does this PR do?
Discussed with @sgugger and @LysandreJik . This PR introduces auto-setup functionality for tutorials and examples (we are sending a parallel PR to accelerate, and maybe Diffusers and Spaces shortly). This allows users to run transformers code, tutorials, and scripts on self-hosted hardware (either their own instance or a cloud instances) including on-demand allocation of the hardware itself on AWS, GCP, Azure, or Lambda Labs, and installation of dependencies. This introduces a similar level of turnkey usage and reproducibility that users only typically expect in Colab, but for any type of hardware on any cloud (we've tested on Paperspace and Coreweave as well, allocating the instance in their UI and then plugging in the IP as a static cluster). Note that Runhouse OSS is facilitating the setup (via SkyPilot) and rpc, but users don't don't need to create a Runhouse account or anything like that, this is strictly inside their own cloud accounts with their own credentials (or using their IP and ssh creds without a cloud account).
Before submitting
Pull Request section?
documentation guidelines, and
here are tips on formatting docstrings.