bittensor  by opentensor

Bittensor SDK for building internet-scale neural networks

Created 5 years ago
1,233 stars

Top 32.0% on SourcePulse

GitHubView on GitHub
Project Summary

Bittensor provides an SDK for building and interacting with an internet-scale, decentralized network of machine intelligence. It enables developers to create "subnets" that offer specialized AI services, rewarding participants (miners and validators) with TAO tokens based on their contributions and performance, as determined by Yuma Consensus.

How It Works

Bittensor operates as a blockchain ("subtensor") that coordinates numerous off-chain "subnets." Each subnet focuses on a specific digital commodity, such as AI model inference or data storage. Subnet validators initiate competitions, and subnet miners respond by providing the best quality commodity. The network incentivizes participation and quality through a token-based reward system, managed by the on-chain Yuma Consensus mechanism.

Quick Start & Requirements

  • Install: python3 -m pip install --upgrade bittensor or via a bash script: /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/opentensor/bittensor/master/scripts/install.sh)"
  • Prerequisites: Python 3.9+ (specifically 3.9 and 3.10 for cubit integration), pip, venv. Windows users require WSL 2. GPU/CUDA not explicitly required for SDK installation but implied for mining/validating.
  • Verification: python3 -m bittensor --version or via Python interpreter: import bittensor as bt; print(bt.__version__).
  • Docs: https://docs.bittensor.com

Highlighted Details

  • Enables creation of decentralized AI marketplaces and services.
  • Utilizes Yuma Consensus for on-chain incentive alignment.
  • Supports various digital commodities beyond AI, like storage and prediction.
  • Offers a comprehensive SDK for subnet development and network interaction.

Maintenance & Community

  • Active development with regular releases.
  • Community support via Discord: https://discord.gg/qasY3HA9F9
  • Project acknowledges contributions from learning-at-home/hivemind.

Licensing & Compatibility

  • License: MIT License.
  • Compatibility: Permissive MIT license allows commercial use and integration with closed-source projects.

Limitations & Caveats

  • Mining and validating operations are not recommended or supported on Windows machines, requiring WSL 2.
  • Specific subnet functionalities may have additional dependencies or requirements not detailed here.
Health Check
Last Commit

18 hours ago

Responsiveness

1 week

Pull Requests (30d)
38
Issues (30d)
12
Star History
24 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.