Einstein-AIT is a decentralized network utilizing the Bittensor protocol for advanced AI/ML model development focused on mathematics, computational thinking, and data analysis. This document guides validators and miners on setup, operation, and contribution.
Einstein-AIT subnet aims to:
- Host a fine-tuning competition for AI/ML models specialized in mathematics and computational reasoning.
- Produce a revenue-generating subscription business through a front-end Math and Data analytics LLM application, leveraging RLFH via real user feedback and response rating.
- Open-source and fine-tune a custom AI/ML model trained on a robust dataset, including a cleaned MIT-IDSS mathematics dataset.
To provide real-world users with an ever-improving AI/ML model capable of computing complex mathematics and data analytics, adapting to the user's education level, and providing appropriate justifications for its results.
- Open Source: Specialized mathematics ML model trained on various datasets, including proprietary data from MIT Institute for Data Systems and Society (IDSS).
- Hugging Face Leaderboard: For fine-tuning the model and allowing miners to compete with custom models.
- Front-End Application: Revenue-generating business model.
- 🚀 Advanced Computational Network: Incentivizing miners to enhance computational resources for complex AI/ML tasks.
- 📈 Performance Commitment: Miners commit to a model type and task volume.
- 💰 Incentive Mechanism:
reward = (0.6 * accuracy_score) + (0.3 * reasoning_score) + (0.1 * time_score)accuracy_scoreis based on the correctness of the result.reasoning_scoremeasures the similarity between the generated text and reference data.time_scoreis derived from the time taken to generate the result.
- 🌟 Continuous Improvement: Introducing new models and features based on demand.
- 💵 Earn as a Validator: Validators earn by sharing computational resources with miners.
- 🛠️ Install dependencies and set up the Einstein-AIT validator node.
- ⚙️ Configure validator settings, including TAO staking.
- 🚀 Start the validator node and process AI/ML tasks from miners.
Detailed instructions for setting up the validator node can be found here.
- 🛠️ Install dependencies and set up the Einstein-AIT miner node.
- 🗂️ Choose a model type and specify task volume.
- 🚀 Start the miner node and contribute computational resources.
Detailed instructions for setting up the miner node can be found in the miners documentations:
We welcome contributions to the Einstein-AIT project! If you have ideas, bug reports, or feature requests, please open an issue on our GitHub repository. To contribute code, fork the repository and submit a pull request.
This repository is licensed under the MIT License. See LICENSE for details.