# Montly Meeting Log

## Wednesday, February 5, 2025
### Attendence: CANCELLED
### Agenda
* Add any updates or requests for help in slack!

### Collaboration Opportunities
*

### Notes
*

## Wednesday, January 8, 2025
### Attendence: Saša Zelenović (Neural Magic,) Simon Mo (Berkley,) Mo McElaney (IBM)

### Agenda
* vLLM website Review meeting with LF happening tomorrow (January 9 at 4pm ET - [staging link](https://vllm.dev-lfprojects5.linuxfoundation.org/) - Simon, Matt, Mo
* runLLM Bot privacy review with Linux Foundation happening today at 2pm ET - Simon, Chenggang, Mo
* vLLM Governance work - Cara, Mo, Matt White
* Neural Magic events/meetups, Blogs, Newsletter - Sasa, Simon
* Planning for PyTorch Conference 2025 - https://events.linuxfoundation.org/pytorch-conference-2025/

### Collaboration Opportunities
* Mo and Cara will reach out to Matt White, at LF who runs the PyTorch Foundation, because he offered to help with vLLM governance stuff - Help vLLM to qualify as a [graduated project](https://lfaidata.foundation/blog/2019/10/10/a-guide-to-hosting-your-project-in-lf-ai/) in the Linux Foundation
* Link to google doc with draft resources we have begun to collect: https://docs.google.com/document/d/1MUvMznhWPIU_Ol-lAwJZQp03O7gmITJ6zaXiBWf09e4/edit?usp=sharing
* Sasa is going to talk to Cara Delia about Red Hat funding for vLLM travel stuff

### Notes

## Wednesday, December 4, 2024
### Attendence: Jonathan Berkhahn (IBM,) Cara Delia (Red Hat,) Saša Zelenović (Neural Magic,) Simon Mo (Berkley,) Mo McElaney (IBM)

### Agenda
* vLLM website Review - [staging link](https://vllm.dev-lfprojects5.linuxfoundation.org/)
* runLLM Bot privacy review with Linux Foundation
* Discuss Project [Incubation](https://lfaidata.foundation/blog/2019/10/10/a-guide-to-hosting-your-project-in-lf-ai/) at LF

### Collaboration Opportunities
* Jonathan Berkhahn is going to work on the developer quick start materials.
* Mo will schedule a meeting with LF so we can talk about the website customization.
* Mo and Cara - Help vLLM to qualify as a [graduated project](https://lfaidata.foundation/blog/2019/10/10/a-guide-to-hosting-your-project-in-lf-ai/) in the Linux Foundation:
* Document current project owners, and current and emeritus committers in OWNERS.md and COMMITTERS.md files
* Document project’s governance (we help projects create a governance model that works for them or simply help them document their existing governance).

### Notes

## Wednesday, November 6, 2024
### Attendence: Simon Mo (Berkley), Saša Zelenović (Neural Magic), Jonathan Berkhan (IBM), Mo McElaney (IBM)

### Agenda
* vLLM website Design
* 2025 Event Plan

### Collaboration Opportunities
* Website design from LF and then Anyscale will make it snazzy
* Matthew Connor and Richard Liaw
* Mo will handle comms with LF to get the website stood up. Will use this [content plan](https://docs.google.com/document/d/1qr08kmNf3o_UyUZjhLT5f6JZa6hWXPiDZbv5AOit7iY/edit?usp=sharing) from the design thinking session we had on August 7.
* Saša is planning an event at Google's office in Boston in February/March
* Blog posts, case studies, things people are doing, etc

### Notes

## Wednesday, September 4, 2024
### Attendence:

### Agenda
* vLLM website paused - waiting on LF AI wordpress
* New public free tier slack! Join link: http://slack.vllm.ai/
* NeurIPS booth ad space with NeuralMagic

### Collaboration Opportunities
* Ray Summit Recap Blog Post - Mo will draft

* Slack Channels Mo has added/changed since this discussion:
* #show-and-tell - make default
* #community-chit-chat - make default
* #code-of-conduct - make default
* #jobs - make default
* #sig-events
* #contributors
* #documentation - make default
* #private-mods
* #github-feed
* #channel-guide
* #private-community-collab
* #security
* Havent yet added: Office Hours channel, thinking NeuralMagic should do that when you are able to add any necessary links. At the very least the channel should have the office hours registration link pinned.

* NeuralMagic is going to be at NeurIPS and offered to give vLLM some advertising space at the booth

### Notes

* vLLM community Slack discussion...
Should the Slack be open to anyone or should it be closed to committors only? Discussed pros and cons. Open means we can attract wider audience, host bigger events, but that would require more moderation work. Should we model after PyTorch where only commiters get access to the community slack and everyone else is pushed to the discussion forums? PyTorch slack application: https://docs.google.com/forms/d/e/1FAIpQLSeADnUNW36fjKjYzyHDOzEB_abKQE9b6gqqW9NXse6O0MWh0A/viewform
* We are bumping this meeting time back by 2 hours moving forward so that we can hopefully get more people from the west coast of the US to attend

## Wednesday, September 4, 2024
### Attendence:

### Agenda
* Promote vLLM track at Ray Summit
* Saša will demo vLLM on Common Room - Do this next month!
* Mo will provide updates on the vLLM website
* Discuss potential code of conduct
* Need email list/marketing tool - Is there an open source tool we can utilize?
* Email the vLLM community - what tool should we use?
* Hubspot
* Ray Summit will invite their attendees as well.

### Collaboration Opportunities
* Ray Summit vLLM track promotion
* Robert has posted on [LI](https://www.linkedin.com/posts/robert-nishihara-b6465444_vllm-raysummit-activity-7229225112200208384-tGiS?utm_source=share&utm_medium=member_desktop) and [X](https://x.com/robertnishihara/status/1823457281406853232) about the vLLM track this year at Ray Summit. Find the [image here](https://drive.google.com/file/d/1ExEW-3uXfh9GkXtdHt9ZFcMXThkk-jpz/view?usp=drive_link) if you want to use that.
* There's a special a promo code, ANYSCALEVLLM15, for the vLLM community. Please kindly share this with your network/channels.
* Who can work with Bradley Gottsman on vLLM Happy Hour registration for the HH Monday, Sept 30th?

### Events in September/October
* Meetup with Invidia
* Presence at PyTorch Conference
* AMD 1-day hackathon
* Meta-Connect
* Ray Summit - AnyScale will create images that we can leverage for posting
* AMD event
* IBM TechXchange

## Wednesday, August 7, 2024
### Attendence: Simon Mo, Sasa Zelenovic, Mo McElaney, Michael Goin

### Agenda
* Design thinking session to do initial gathering of requirements for a new vLLM Website!
* Submit Project Overview and Panelist [sessions for Ray Summit](https://raysummit.anyscale.com/flow/anyscale/raysummit2024/callforpapers/form/session)
* Follow up on community metrics tracking (orbit/common room)
* Next meeting Wednesday, September 4

### Collaboration Opportunities
* NM has a Common Room account and got approval from Simon to plug in the vLLM GitHub to see if it's useful
* Short/Lite design session to confirm the initial sitemap for the vLLM website
* [View mural](https://app.mural.co/t/ibmopentech1494/m/ibmopentech1494/1723041158118/f12659a30f231eb6bd4e04b17eb2a608e2db1f1d?sender=u336dfa5b461079c2e09e8775) for more info, email [e-mail] for edit access
* Awaiting confirmation on designer from Red Hat to help us build the website
* Mo McElaney will assist with pulling content over from GitHub, etc
* Finalized panel proposal for Ray Summit

## Thursday 11 July, 2024
### Attendence: Mo McElaney (IBM), Sasa Zelenovic (Neural Magic), Michael Goin (Neural Magic), Cody Yu (AnyScale), Woosuk Kwon (Berkley), Nick Hill (IBM), Simon Mo (Berkley), Nicole Kim (Neural Magic)

### Agenda
* Roundtable introductions
* Review existing community growth initiatives

### Collaboration Opportunities
* Backlog of topics for blogs/workshop goals
* Beginner materials
* Onboarding materials
* Neural Magic has a content roadmap, 6 blog posts in their backlog
* llm-compressor simple examples
* Llama-70B quantized
* Marlin Fp8
* W8A8 is now supported in vLLM
* Marlin 2:4
* KServe example
* Website
* Is there budget for this? There is budget but it is reserved for compute
* No pattern library or styleguide for the project yet
* vllm.ai domain is live
* Mo will send an invite to Nicole/Sasa to discuss possibilities around a website

* Community metrics tracking
* NM is tracking github metrics/project health
* Orbit - vLLM has an account but they don't love it
* NM has a Common Room account and is willing to plug it in to vLLM and see if it's useful

* Ray Summit
* Mo will talk to Bradley about scheduling a general session on vLLM updates and another session with a panel (users and adopters or contributors)
* The purpose is open q&a with conference attendees, motivate folks to contribute.

### Planned as of 07-11
* Biweekly OH from NM
* Upcoming: Deep dive into Multi-Modal (August)
* Blog post from NM about FP8 (big launch next week)
* Meetup in SFO28 AWS office July 24
* Pipeline Parallelism
* Crusoe Hackathon: NM and Anyscale presenting FP8 on vLLM
* vLLM session at virtual IBM TechXchange Conference July 16
* Ray Summit vLLM track - September 30

### Existing Assets
* Project Docs
* Blog: anyscale and nm
* Intermediate level Talks
* general vLLM talks (previous meetups)
* deep dive spec decode
* deep dive in hardware support