OpenGradient is excited to announce our first Modelthon, co-sponsored by Allora to begin the year with a BANG!
Submit ML forecasting models to earn prizes, recognition, and acquire general awesomeness.
OpenGradient provides a decentralized on-chain ecosystem for AI and ML inference. Our Modelthon invites participants to develop and submit Spot Forecasting Models that will operate and compete on live data, or unspecified freestyle Web3 models for our or your own data.
There will be winners for each of the following competitions
- BTC Spot Forecast - Train a model to predict 1hr spot BTC returns with our OHLC dataset.
- ETH Spot Forecast - Train a model to predict 1hr spot ETH returns with our OHLC dataset.
- Freestyle - Train and submit your own model for any Web3 use case!
Cash Prizes + Recognition
The total cash prize pool is $50,000
- $20,000 allocated to ETH spot forecast winners
- $20,000 allocated to BTC spot forecast winners
- $10,000 allocated to freestyle winners
All winners will be announced and featured on OpenGradient X (224K+ followers) and Discord (183K+ members).
Judging Criteria
Judges from OpenGradient and Allora have deep experience in AI/ML and blockchain. Judging criteria for each track is listed below:
- Spot Forecasts: A score will be given where 85% of the score comes from an objective similarity measure (more details in the post-registration email) and 15% for completeness of your model description.
- Freestyle: A holistic score will be given by expert competition judges using, among other things, utility, ingenuity, design, description, results, etc.
Register to Participate
- Register HERE now, no later than Feb 1.
- You will receive an email within a few days with further instructions and links to necessary documents and data. Be sure to use the same email you used to sign up and connect with the Model Hub for uploading your models.
- Make sure the email you entered on the form is the same as the email you use to log into the Model Hub.
- Join the #modelthon-hq and #modelthon-support channels on the official OpenGradient Discord for live support and interacting with the OpenGradient team and other participants.
Getting Started
- Must register first!
- Download the OpenGradient training data set (instructions in post-registration email).
- Build a model using sklearn, tensorflow, pytorch, random numbers, or other tools.
- Turn the model into ONNX format (help for this will be in post-registration email).
- Submit your model to the OpenGradient Model Hub (see Model Submission below).
Model Submission
- Install the OpenGradient SDK.
- Upload your model via SDK or OpenGradient Model Hub.
- For spot forecast models, confirm your model works using the OG Modelthon Test Package (instructions in the post-registration email).
- Make sure you include a full description of your model on OpenGradient Model Hub for full points - graphics and plots make it even cooler.
- Submit your model info on the Modelthon Submission page (more info will be provided in the post-registration email).
Important Dates
- Model and submission page must be submitted before Feb 4 00:00 UTC.
- On Feb 4 at 00:00:00 UTC we will begin running everyone’s model and you can track all models’ progress live on the Modelthon Dashboard!
- On Feb 21 at 00:00 UTC, we will stop running models and declare winners within 48 hours!
Don’t miss out! Register now and start building!
About OpenGradient
OpenGradient is a leading decentralized platform unlocking scalable AI compute on-chain, empowering developers to build secure, intelligent, and optimized decentralized applications. We provide on-chain AI model hosting, permissionless composability, secure inference execution, and seamless access to accelerate open-source AI and foster next-gen applications. Headquartered in New York City, the team comprises global talent with deep expertise in AI/ML, blockchain, crypto, and Web3 ecosystem. Learn more at https://opengradient.ai and follow us on X and LinkedIn.
About Allora Network
Allora is a self-improving decentralized AI network. Allora enables applications to leverage smarter, more secure AI through a self-improving network of ML models. By combining innovations in crowdsourced intelligence, reinforcement learning, and regret minimization, Allora unlocks a vast new design space of applications at the intersection of crypto and AI. To learn more about Allora Network, visit the Allora website, X, Blog, Discord, Telegram, Research Hub, and Developer Docs.