Skip to content

opentensor/bittensor

Repository files navigation

Bittensor SDK

Discord Chat PyPI version License: MIT


Internet-scale Neural Networks

DiscordNetworkResearch


Overview of Bittensor

Welcome! Bittensor is an open source platform on which you can produce competitive digital commodities. These digital commodities can be machine intelligence, storage space, compute power, protein folding, financial markets prediction, and many more. You are rewarded in TAO when you produce best digital commodities.

The Bittensor SDK

The Opentensor Foundation (OTF) provides all the open source tools, including this Bittensor SDK, the codebase and the documentation, with step-by-step tutorials and guides, to enable you to participate in the Bittensor ecosystem.

This Bittensor SDK contains ready-to-use Python packages for interacting with the Bittensor ecosystem, writing subnet incentive mechanisms, subnet miners, subnet validators and querying the subtensor (the blockchain part of the Bittensor network).


Is Bittensor a blockchain or an AI platform?

In Bittensor there is one blockchain, and many platforms that are connected to this one blockchain. We call these platforms as subnets, and this one blockchain subtensor. So, a subnet can be AI-related or it can be something else. The Bittensor network has a number of distinct subnets. All these subnets interact with subtensor blockchain. If you are thinking, "So, subnets are not part of the blockchain but only interact with it?" then the answer is "yes, exactly."

Subnets

Each category of the digital commodity is produced in a distinct subnet. Applications are built on these specific subnets. End-users of these applications would be served by these applications.

Subnet validators and subnet miners

Subnets, which exist outside the blockchain and are connected to it, are off-chain competitions where only the best producers are rewarded. A subnet consists of off-chain subnet validators who initiate the competition for a specific digital commodity, and off-chain subnet miners who compete and respond by producing the best quality digital commodity.

Yuma Consensus

Scores are assigned to the top-performing subnet miners and subnet validators. The on-chain Yuma Consensus determines the TAO rewards for these top performers. The Bittensor blockchain, the subtensor, runs on decentralized validation nodes, just like any blockchain.

This SDK repo is for Bittensor platform only This Bittensor SDK codebase is for the Bittensor platform only, designed to help developers create subnets and build tools on Bittensor. For subnets and applications, refer to subnet-specific websites, which are maintained by subnet owners.

Release Notes

See Bittensor SDK Release Notes.


Install Bittensor SDK

Before you can start developing, you must install Bittensor SDK and then create Bittensor wallet.

Upgrade

If you already installed Bittensor SDK, make sure you upgrade to the latest version. Run the below command:

python3 -m pip install --upgrade bittensor

Install on macOS and Linux

You can install Bittensor SDK on your local machine in either of the following ways. Make sure you verify your installation after you install:

Install using a Bash command

This is the most straightforward method. It is recommended for a beginner as it will pre-install requirements like Python, if they are not already present on your machine. Copy and paste the following bash command into your terminal:

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/opentensor/bittensor/master/scripts/install.sh)"

For Ubuntu-Linux users If you are using Ubuntu-Linux, the script will prompt for sudo access to install all required apt-get packages.

Install using pip3 install

python3 -m venv bt_venv
source bt_venv/bin/activate
pip install bittensor

Install from source

  1. Create and activate a virtual environment

  2. Clone the Bittensor SDK repo

git clone https://github.com/opentensor/bittensor.git
  1. Install

You can install using any of the below options:

  • Install SDK: Run the below command to install Bittensor SDK in the above virtual environment. This will also install btcli.

    pip install bittensor
  • Install SDK with torch: Install Bittensor SDK with torch.

     pip install bittensor[torch]

    In some environments the above command may fail, in which case run the command with added quotes as shown below:

      pip install "bittensor[torch]"
  • Install SDK with cubit: Install Bittensor SDK with cubit.

    1. Install cubit first. See the Install section. Only Python 3.9 and 3.10 versions are supported.
    2. Then install SDK with pip install bittensor.

Install on Windows

To install and run Bittensor SDK on Windows you must install WSL 2 (Windows Subsystem for Linux) on Windows and select Ubuntu Linux distribution.

After you installed the above, follow the same installation steps described above in Install on macOS and Linux.

ALERT: Limited support on Windows While wallet transactions like delegating, transfer, registering, staking can be performed on a Windows machine using WSL 2, the mining and validating operations are not recommended and are not supported on Windows machines.


Verify the installation

You can verify your installation in either of the below ways:

Verify using btsdk version

python3 -m bittensor

The above command will show you the version of the btsdk you just installed.

Verify using Python interpreter

  1. Launch the Python interpreter on your terminal.

    python3
  2. Enter the following two lines in the Python interpreter.

    import bittensor as bt
    print( bt.__version__ )

    The Python interpreter output will look like below:

    Python 3.11.6 (main, Oct  2 2023, 13:45:54) [Clang 15.0.0 (clang-1500.0.40.1)] on darwin
    Type "help", "copyright", "credits" or "license" for more information.
    >>> import bittensor as bt
    >>> print( bt.__version__ )
    <version number>

You will see the version number you installed in place of <version number>.

Verify by listing axon information

You can also verify the Bittensor SDK installation by listing the axon information for the neurons. Enter the following lines in the Python interpreter.

import bittensor
metagraph = bittensor.Metagraph(1)
metagraph.axons[:10]

The Python interpreter output will look like below.

[AxonInfo( /ipv4/3.139.80.241:11055, 5GqDsK6SAPyQtG243hbaKTsoeumjQQLhUu8GyrXikPTmxjn7, 5D7u5BTqF3j1XHnizp9oR67GFRr8fBEFhbdnuVQEx91vpfB5, 600 ), AxonInfo( /ipv4/8.222.132.190:5108, 5CwqDkDt1uk2Bngvf8avrapUshGmiUvYZjYa7bfA9Gv9kn1i, 5HQ9eTDorvovKTxBc9RUD22FZHZzpy1KRfaxCnRsT9QhuvR6, 600 ), AxonInfo( /ipv4/34.90.71.181:8091, 5HEo565WAy4Dbq3Sv271SAi7syBSofyfhhwRNjFNSM2gP9M2, 5ChuGqW2cxc5AZJ29z6vyTkTncg75L9ovfp8QN8eB8niSD75, 601 ), AxonInfo( /ipv4/64.247.206.79:8091, 5HK5tp6t2S59DywmHRWPBVJeJ86T61KjurYqeooqj8sREpeN, 5E7W9QXNoW7se7B11vWRMKRCSWkkAu9EYotG5Ci2f9cqV8jn, 601 ), AxonInfo( /ipv4/51.91.30.166:40203, 5EXYcaCdnvnMZbozeknFWbj6aKXojfBi9jUpJYHea68j4q1a, 5CsxoeDvWsQFZJnDCyzxaNKgA8pBJGUJyE1DThH8xU25qUMg, 601 ), AxonInfo( /ipv4/149.137.225.62:8091, 5F4tQyWrhfGVcNhoqeiNsR6KjD4wMZ2kfhLj4oHYuyHbZAc3, 5Ccmf1dJKzGtXX7h17eN72MVMRsFwvYjPVmkXPUaapczECf6, 600 ), AxonInfo( /ipv4/38.147.83.11:8091, 5Hddm3iBFD2GLT5ik7LZnT3XJUnRnN8PoeCFgGQgawUVKNm8, 5DCQw11aUW7bozAKkB8tB5bHqAjiu4F6mVLZBdgJnk8dzUoV, 610 ), AxonInfo( /ipv4/38.147.83.30:41422, 5HNQURvmjjYhTSksi8Wfsw676b4owGwfLR2BFAQzG7H3HhYf, 5EZUTdAbXyLmrs3oiPvfCM19nG6oRs4X7zpgxG5oL1iK4MAh, 610 ), AxonInfo( /ipv4/54.227.25.215:10022, 5DxrZuW8kmkZPKGKp1RBVovaP5zHtPLDHYc5Yu82Z1fWqK5u, 5FhXUSmSZ2ec7ozRSA8Bg3ywmGwrjoLLzsXjNcwmZme2GcSC, 601 ), AxonInfo( /ipv4/52.8.243.76:40033, 5EnZN591jjsKKbt3yBtfGKWHxhxRH9cJonqTKRT5yTRUyNon, 5ChzhHyGmWwEdHjuvAxoUifHEZ6xpUjR67fDd4a42UrPysyB, 601 )]
>>>

Release Guidelines

Instructions for the release manager: RELEASE_GUIDELINES.md document.

Contributions

Ready to contribute? Read the contributing guide before making a pull request.

License

The MIT License (MIT) Copyright © 2024 The Opentensor Foundation

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Acknowledgments

learning-at-home/hivemind