Ensure you're running Subtensor locally to minimize outages and improve performance.
Refer to the Run a Subtensor Node Locally guide.
Check the minimum compute requirements for our subnet, defined in the Minimum compute YAML configuration.
Clone the repository and navigate to the project directory:
git clone https://github.com/natixnetwork/natix-subnet.git && cd natix-subnetPython 3.11 is required. Create a virtual environment and install validator dependencies:
python3.11 -m venv venv
source venv/bin/activate
pip install -r requirements.validator.txtTo also run synthetic image generation (optional background process):
pip install -r requirements.validator-full.txtbtcli s register --netuid 72 --wallet.name [wallet_name] --wallet.hotkey [wallet.hotkey] --subtensor.network finneybtcli s register --netuid 323 --wallet.name [wallet_name] --wallet.hotkey [wallet.hotkey] --subtensor.network testUpdate your validator.env file with your configuration:
NETUID=72
SUBTENSOR_NETWORK=finney
SUBTENSOR_CHAIN_ENDPOINT=wss://entrypoint-finney.opentensor.ai:443
WALLET_NAME=default
WALLET_HOTKEY=default
VALIDATOR_AXON_PORT=8092
VALIDATOR_PROXY_PORT=10913
PROXY_CLIENT_URL=https://hydra.natix.network
DEVICE=cuda
WANDB_API_KEY=your_wandb_api_key_here
HUGGING_FACE_TOKEN=your_hugging_face_token_hereTo run the validator:
pm2 start scripts/ecosystem.validator.config.jsOptional flags:
--no-auto-updates: Disables automatic code updates--no-self-heal: Disables automatic restart every 6 hours
Please note that you need to expose the port numbers you define by VALIDATOR_AXON_PORT and VALIDATOR_PROXY_PORT for incoming requests.
That’s it — you’re ready to validate!