🍃 Lilypad Project Report: November 07, 2024
Lilypad v2.9.0 release, Updated Lilybit Rewards calculation, Community Growth, and more!
🌎 Overview
Welcome to the latest Lilypad Project Report! We're excited to bring you updates on the strides we’re making to democratize access to GPU compute and build a robust, decentralized ecosystem.
The Lilypad Network aims to democratize access to GPU compute by aggregating idle GPU capacity and providing a market-matching system for end users. Currently in Testnet, our IncentiveNet program rewards community members for supporting the network. Through Lilybit_ rewards, you can earn by running a Lilypad node, making open-source contributions, and joining our growing Ambassador program, and soon, by participating in educational initiatives.
The goals of IncentiveNet are to test the network orchestration architecture, proof various game theoretic mechanisms of the system, and provide a production grade offering for both GPU compute supply and compute demand. Over 2,200 nodes have been on-boarded to date running thousands of jobs from Stable diffusion 1.0 to Ollama.
Leaderboard Highlights
The Resource Provider (RP) Leaderboard and RP Points Dashboard track Lilybit earnings, giving RPs a clear view of their contributions.
Our Community Dashboard showcases rewards for open-source contributions and Ambassadors – making it a one-stop hub for all participants to see their impact!
The road to Mainnet
As we move closer to Mainnet, IncentiveNet is at the heart of Lilypad’s journey, enabling us to fine-tune our network economics and ensure reliable, scalable infrastructure. To our early Resource Providers and ecosystem partners – thank you for helping us shape the future of decentralized AI.
Ready to join us? Run a node, explore the potential of Lilypad, and be part of a more distributed, accessible AI ecosystem!
⚒️ Engineering Update
The Lilypad Engineering team ran a group offsite in Austin, USA with the goal of accelerating our trajectory towards Mainnet launch. The team workshopped different approaches to network orchestration and the job module flow, dove into learnings from the IncentiveNet, and strengthened the Lilypad development pipeline in order to launch Mainnet according to plan!
As IncentiveNet scales, our team has become familiar with managing the Lilypad economy at scale. Thousands of Resource Offers (GPUs offering to run jobs) must be accounted for in the Lilypad Solver (market matching system) and matched with compute job demand. The job creator and management workflow must ensure the right amount of information is provided without compromising job privacy. In addition, our team is evaluating many Blockchain based approaches, including the current use of Arbitrum Sepolia Testnet in IncentiveNet, in order to ensure the Mainnet can scale quickly and cost effectively.
Update RPs to the latest version! - v2.9.0 Einstein
"Imagination is more important than knowledge." ~ Albert Einstein
Lilypad v2.9.0 - Einstein was released, requiring Lilypad RPs to update!
The aim of this update (in addition to the past few updates) has been focused on consistency and smooth scaling of the job matching system.
To fairly reward GPUs optimized for AI jobs, Lilybits will be provided based on GPU card info (VRAM, Tensor cores, Cuda cores) in addition to the PoW uptime multiplier. The calculation will no longer use hashrate, but RPs will still be required to run PoWs. Read more.
RPs will report their Lilypad version to the network
This release includes a docker-compose file for RPs to easily get setup and running!
Updated Lilybit_ Rewards Calculation
GPUs that are optimized for running AI models are not particularly strong at running Proof of Work calculations. Thus, to more correctly reward GPUs like a100s and h100s, the Lilybit rewards calculation was updated to utilise GPU card info in addition to the daily rewards multipler to determine points.
Base Lilybit_ points are now assigned by the GPU card type using three parameters: VRAM, Cuda core count, and Tensor core count. RP hashrate will no longer be used, but PoWs are still required to determine the RP uptime multiplier.
We are aware that some RPs may have cheated the system and have multiple data points for this. There will be an audit on cheating before the final allocation of rewards (details will be announced closer to the time). Lilypad reserves the right to alter points allocations based on activity suspected of breaking the spirit of the IncentiveNet Program, not operating in good faith or being deliberately malicious.
Change to slashing
The Lilypad IncentiveNet employs a slashing mechanism to penalize RPs that are not running consistently on the network. After listening to community feedback, the slashing mechanism now uses a tiered/progressive system to more fairly slash RPs. Penalties will increase progressively the longer the node remains online. 1-5 days offline will result in a 2.5% slash per day, 5-10 days offline a 5% slash per day, and more than 10 days offline a 10% slash per day.
In addition, RPs will earn 2 days of a “grace period” after every 30 days of continuous service provided. These 2 days will be applied to 2 subsequent down days recorded by the RP, allowing the RP to avoid slashing for these 2 days.
Grace Period days do not accumulate to more than 2 days ever. Once used, the 30 day count to obtain the 2 days restarts.
Stay connected on the Lilypad Discord and follow along on X!
⚡ Developer Events & DX Updates
Lilypad Ambassadors
Since launching in July, the Lilypad Ambassadors Program has seen an incredible response from our community! We couldn’t be prouder of this first group of ambassadors, who’ve been instrumental in building and supporting Lilypad Network. Their enthusiasm and contributions continue to drive our mission forward.
A special shoutout to Lilypad Ambassador lead Kabularasa, who developed a Telegram bot to notify the operator of RP downtime. This proactive contribution helps keep our community informed and engaged, embodying the true spirit of decentralized collaboration.
Lilypad and Titan Network X Space
Lilypad and Titan Network hosted a X Space to explore why decentralized compute is foundational to an open web and AI framework, focusing on how we can drive adoption in this emerging field. The discussion covered the critical role of decentralized infrastructure in powering the future of AI, emphasizing scalability and accessibility.
Missed the conversation? Listen to the replay for more insights 🎧
Exploring the Future of Decentralized AI with Lilypad and Parasail
"Scalability through market forces." – Lilypad Founder @DeveloperAlly
In an insightful X/Twitter Space, Lilypad and Parasail Network dove discussed their thoughts on the future of decentralized AI. The conversation centered around how marketplace dynamics empower decentralized compute networks like Lilypad to scale naturally. As demand for compute rises, more GPUs join the network, allowing Lilypad to grow and evolve in real-time. This is the true power of Web3 creating networks that expand with demand, driven by community and participation.
Missed the live session? Listen in 🎧
Weekly AMA Sessions – Connect with the Lilypad Team!
Curious about Lilypad's latest updates, projects, and vision? Join us every Wednesday for our Weekly AMA Sessions! These sessions are your chance to ask questions, gain insights, and interact directly with the Lilypad team. We’re excited to share our journey with you and dive into topics that truly matter to our community.
Missed Last Week’s AMA?
Catch up with DeveloperSteve and Taosheng as they answered community questions and shared exclusive updates.
Flexible Times: To make these sessions accessible to everyone, we alternate the timing each week, ensuring our global community of RPs, developers, and enthusiasts can join in.
🗓️ Find a session that works best for you on the Lilypad Calendar
Special Feature – Chinese AMA is coming soon!
We’re hosting a dedicated session for our Chinese community led by Taosheng. Stay tuned for details and join the conversation.
🌟Community Spotlight
New on the Blog Unfolding the Mysteries of Protein Structure
Our latest blog post explores the fascinating world of Protein Structure through ColabFold on the Lilypad Network. Dive into how our network supports BioML jobs, giving the BioML community accessible compute power to run, fine-tune models, and eventually earn attribution through Lilypad Modules.
Tadpole Talent Board – Dive into a pond of AI and Blockchain Opportunities
Looking to make your mark in the fast-paced world of AI and Blockchain?
Join our Tadpole Talent Board on Discord, where you can showcase your skills, connect with potential collaborators, and explore exciting roles within the Lilypad ecosystem. Show off what you work on and how you would like to contribute!
🔮 What's Next?
As we push towards Mainnet launch, the Lilypad core team is heads down building out and testing the various components of the Lilypad economy. Updates continue to be pushed to the IncentiveNet further informing development choices for Mainnet. Feedback from the RP community and insights from the end user ecosystem of ML researchers, labs, and enterprises has been critical to the Lilypad Whitepaper, Tokenomics, and network design!
Stay updated on all the latest Resource Provider developments in the updates-rp channel on Discord!
☎️ Contact Us
💬 Got questions or need guidance? Join our next AMA, chat with us on Discord, or follow us on X for the latest news and updates. We’re here to support our community every step of the way!