Omron SN2 (𝜏, β)

Proof of Inference

Documentation • X • Inference Labs • Stats • Studio • Dashboard

Omron represents a significant stride in enhancing the Bittensor network, aiming to create the world's largest peer-to-peer Verified Intelligence network, by building a Proof-of-Inference system for the Bittensor network. This initiative aligns with the Opentensor foundation's criteria for innovative subnet solutions. zk-ML allows AI models to be converted into a unique 'fingerprint,' a circuit that can be used to verify that a model's prediction was generated by a specific AI model, thereby providing what we term as Proof-of-Inference.

Miners and Validators Functionality

Incentive Mechanism and Reward Structure

Omron incentivizes miners and validators on Subnet 2 to contribute to the generation and validation of high-quality, secure, and efficient verified AI predictions using a specialized reward mechanism that aligns with the unique aspects of zero-knowledge machine learning (zk-ML) and decentralized AI. Currently Zero-knowledge proofs are generally more CPU computationally intensive and opens the opportunity for non-GPU miners to participate however the end goal is to further incentivize the development of proving systems optimized for GPU based operations. The incentives are based around Miners creating succinct and efficient models which can be circuitized with a zero-knowledge proving system.

The reward mechanism for Subnet 2 scores the initial AI predictions based on the cryptographic integrity and time to generate zk-proofs along with the outputs, rather than solely on end results. This approach reduces the computational burden on validators as zk-proofs confirm the source model and integrity of AI predictions efficiently.

Miners

• Receive input data from validators on the subnet.
• Generate predictions using custom, verifiable AI models that have been converted into zero knowledge circuits
• Return the generated content to the requesting validator for validation and distribution.

Validators

• Produce input data and distribute requests for verified inference throughout miners participating on the subnet
• Confirm that miners are acting faithfully, by verifying the authenticity of the miner's returned zero knowledge proof
• Score results from miners based on performance metrics such as proof size and response time

Quickstart

Run the below command to install Omron and it's dependencies.

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/inference-labs-inc/omron-subnet/main/setup.sh)"

See full setup guide →

Register on the SN

btcli subnet register --subtensor.network finney --netuid 2 --wallet.name {your_coldkey} --wallet.hotkey {your_hotkey}

Run the miner

Docker Instructions (not supported during the competition running from Feb 6 - Apr 20)

#### With docker compose (recommended)

---
services:

  omron-miner:
    image: ghcr.io/inference-labs-inc/omron:latest
    restart: unless-stopped
    ports:
      - 8091:8091
    volumes:
      # Update this path to your .bittensor directory
      # Note: use /root/.bittensor instead of /home/ubuntu/.bittensor if you set PUID to 0
      - {path_to_your_.bittensor_directory}:/home/ubuntu/.bittensor
    environment:
      # This UID needs to be able to read/write to your .bittensor directory, either update the UID or the directory permissions
      - PUID=1000
    labels:
      - com.centurylinklabs.watchtower.enable=true  # Enables Watchtower for this container
    command: miner.py --wallet.name {your_miner_key_name} --wallet.hotkey {your_miner_hotkey_name} --netuid 2

  # Use Watchtower automatically update containers
  watchtower:
    image: containrrr/watchtower:latest
    restart: unless-stopped
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock:ro
    command: --interval 60 --cleanup --label-enable

With docker cli

docker run -d \
  --name omron-miner \
  -p 8091:8091 \
  -v {path_to_your_.bittensor_directory}:/home/ubuntu/.bittensor \
  -e PUID=1000 \
  --restart unless-stopped \
  ghcr.io/inference-labs-inc/omron:latest \
  miner.py \
  --wallet.name {your_miner_key_name} \
  --wallet.hotkey {your_miner_hotkey_name} \
  --netuid 2

With pm2

Important

Ensure you are within the ./neurons directory before using the commands below to start your miner

cd neurons

pm2 start miner.py --name miner --interpreter ../.venv/bin/python --kill-timeout 3000 -- \
--netuid 2 \
--wallet.name {your_miner_key_name} \
--wallet.hotkey {your_miner_hotkey_name}

Or run this command with make pm2-miner WALLET_NAME={your_miner_key_name} HOTKEY_NAME={your_miner_hotkey_name}

Run the validator

Docker Instructions (not supported during the competition running from Feb 6 - Apr 20)

#### With docker compose (recommended)

---
services:

  omron-validator:
    image: ghcr.io/inference-labs-inc/omron:latest
    restart: unless-stopped
    ports:
      - 8443:8443
      - 9090:9090  # In case you use prometheus monitoring
    volumes:
      # Update this path to your .bittensor directory
      # Note: use /root/.bittensor instead of /home/ubuntu/.bittensor if you set PUID to 0
      - {path_to_your_.bittensor_directory}:/home/ubuntu/.bittensor
    environment:
      # This UID needs to be able to read/write to your .bittensor directory, either update the UID or the directory permissions
      - PUID=1000
    labels:
      - com.centurylinklabs.watchtower.enable=true  # Enables Watchtower for this container
    command: validator.py --wallet.name {validator_key_name} --wallet.hotkey {validator_hot_key_name} --netuid 2

  # Use Watchtower automatically update containers
  watchtower:
    image: containrrr/watchtower:latest
    restart: unless-stopped
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock:ro
    command: --interval 60 --cleanup --label-enable

With docker cli

docker run -d \
  --name omron-validator \
  -p 8443:8443 \
  -p 9090:9090 \
  -v {path_to_your_.bittensor_directory}:/home/ubuntu/.bittensor \
  -e PUID=1000 \
  --restart unless-stopped \
  ghcr.io/inference-labs-inc/omron:latest \
  validator.py \
  --wallet.name {validator_key_name} \
  --wallet.hotkey {validator_hot_key_name} \
  --netuid 2

With pm2

Important

Ensure you are within the ./neurons directory before using the commands below to start your validator

cd neurons

pm2 start validator.py --name validator --interpreter ../.venv/bin/python --kill-timeout 3000 -- \
--netuid 2 \
--wallet.name {your_validator_key_name} \
--wallet.hotkey {your_validator_hotkey_name}

Or run this command with make pm2-validator WALLET_NAME={validator_key_name} HOTKEY_NAME={validator_hot_key_name}

Miner

Miners contribute to this subnet by providing compute to generate output from, and prove AI model inferences. Miners receive workloads from validators in the form of input data, perform verified inferences on those inputs and respond with output along with a zero knowledge proof of inference.

Important

As of February 2025, the miner should be run on a bare-metal MacOS machine with support for Metal GPU acceleration, to optimize for performance.

Hardware requirements

Minimum

Component	Requirement
CPU	8 core 3.2GHz
RAM	32GB
Network Up	400Mbps
Network Down	400Mbps
Storage	100GB

Recommended

Note

Exceeding these requirements in terms of storage, network and CPU speed will most likely result in higher rewards due to performance incentivization.

Component	Recommendation
CPU	8 core 3.6GHz
RAM	64GB
Network Up	1Gbps
Network Down	1Gbps
Storage	400GB
Storage Medium	SSD

Validator

Validators are responsible for verifying model outputs as provided by miners, and updating that miner's score based on the verification results.

Hardware requirements

Important

As of February 2025, the validator must be run on a bare-metal MacOS machine with support for Metal GPU acceleration.

Though AWS instances for metal are recommended, any MacOS machine with a Metal GPU is sufficient.

Minimum

Component	Requirement
Instance	mac2-m2pro.metal (AWS)
CPU	Apple M2 Pro (12-core)
RAM	32GB
Network Up	10Gbps
Network Down	10Gbps
Storage	2TB SSD

Recommended

Component	Recommendation
Instance	mac2-m1ultra.metal (AWS)
CPU	Apple M1 Ultra (20-core)
RAM	128GB
Network Up	10Gbps
Network Down	10Gbps
Storage	2TB+ SSD