Loading...
Loading
Overview
Registration Activity
Neuron registration trends
This Interval
0
— burn · — PoW
Target / Interval
Loading
Neuron registration trends
Ecommerce companies struggle to deliver personalized product recommendations at scale. Building and testing recommendation prompts requires expensive infrastructure, manual evaluation cycles, and months of iteration to find what actually drives conversions. Teams either rely on generic algorithms that underperform or invest heavily in proprietary systems that don't improve.
Bitrecs V2 enables distributed prompt optimization by allowing teams to submit artifact specifications (prompt templates, model parameters, and sampling settings) for evaluation against real ecommerce tasks. A network of validators automatically tests each submission against diverse product recommendation scenarios, measures performance using epsilon-Pareto dominance scoring, and ranks results. Top-performing prompts receive rewards, creating incentives for continuous improvement without requiring participants to maintain evaluation infrastructure.
On the Bitrecs V2 subnet, miners submit optimized artifact.yaml files via GitHub gist, and validators run them through multi-stage screening and comprehensive evaluation pipelines. Submissions are scored on performance across multiple ecommerce environments, with weights set onchain for the highest-performing artifacts to receive emissions.
Use cases were auto-generated from the GitHub repository and may not reflect the latest changes.
Bitrecs V2 is a prompt evolution subnet that rewards miners for optimizing artifact.yaml files (containing prompts, models, and parameters) against a rotating set of ecommerce evaluation tasks. Miners submit artifacts via CLI and are scored using a winner-take-all mechanism based on performance across multiple environments, with validators running evaluations and setting weights onchain for top performers.
Bitrecs V2 separates inference from prompt evolution. Miners submit optimized artifact.yaml files via GitHub gist to the CLI, which creates onchain commitments. Artifacts contain prompts, model selection, and sampling parameters targeting ecommerce recommendation tasks.
Validators run a two-stage evaluation: screeners verify data consistency, then validators evaluate against rotating task sets using spawned Docker containers running the bitrecs-evals image. Validators compute WTA scores, set weights onchain, and emit rewards to top performers. Optional BYOK support allows miners to provide their own OpenRouter management keys for inference.
Artifacts are evaluated through a 2-stage screening process followed by validator evaluation. Screeners run basic consistency checks; validators then evaluate against rotating ecommerce tasks using multiple environments. The WTA scoring engine applies epsilon tolerances for robustness, accounts for sample sizes, and uses linear decay (3-day grace period, 5% daily reduction to 25% floor).
Democratizing Bittensor Mining.
This information was auto-extracted from the GitHub repository and may be incomplete or out of date. Always refer to the subnet's GitHub repository for the latest instructions.
Only the top-performing miner receives significant emissions based on surpassing earlier participant thresholds.
The sections above were auto-generated from the GitHub repository and may not reflect the latest changes. View repository
FAQs were auto-generated from the GitHub repository and may not reflect the latest changes.
Holder & total-staked metrics as of 1h ago (hourly snapshot)
Miners only submit YAML artifacts; no heavy computation required locally. Evaluation is performed by validators.
Validators run Docker containers for evaluation; Docker daemon required. Child containers spawned during evaluation runs.
Install UV package manager
curl -LsSf https://astral.sh/uv/install.sh | sh
Clone the Bitrecs V2 repository
git clone [email protected]:bitrecs/bitrecs-v2.git
Sync and setup environment with UV
uv sync
Configure environment variables for subnet connection and optional BYOK setup
Create .env file with SUBTENSOR_ADDRESS=wss://test.finney.opentensor.ai:443, SUBTENSOR_NETWORK=test, and OPENROUTER_MGMT_KEY if using OPEN_ROUTER provider
Create and configure artifact submission file
Create artifact.yaml with required fields: name, version_num (must be '1'), status (must be 'screening_1'), miner_hotkey, provider (CHUTES or OPEN_ROUTER), model, system_prompt_template, user_prompt_template, sampling_params
Host artifact on GitHub gist for submission
Upload artifact.yaml to https://gist.github.com/ (single commit, <24 hours old)
Submit artifact for evaluation (takes up to 1 minute)
uv run bitrecs_cli.py upload --github-account mygithubaccount --gist-id your_gist_id --coldkey-name default --hotkey-name default
Update system packages
sudo apt-get update && sudo apt-get upgrade
Install Docker
sudo apt install apt-transport-https ca-certificates curl software-properties-common && curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg && echo 'deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable' | sudo tee /etc/apt/sources.list.d/docker.list > /dev/null && sudo apt update && sudo apt install docker-ce
Install UV package manager
curl -LsSf https://astral.sh/uv/install.sh | sh
Create working directory (must be named 'bitrecs' for docker-compose)
mkdir bitrecs && cd bitrecs
Setup wallet with coldkey and hotkey
uv init && uv pip install bittensor-cli && uv run btcli w regen-coldkeypub --ss58 COLDKEY_ADDR && uv run btcli w regen-hotkey
Pull required Docker images
docker pull ghcr.io/bitrecs/bitrecs-v2:main && docker pull ghcr.io/bitrecs/bitrecs-evals:main
Create and configure environment file with all required variables
touch .env and configure with DEBUG, NETUID, BITRECS_PLATFORM_URL, BITRECS_PLATFORM_API_KEY, SUBTENSOR_NETWORK, SUBTENSOR_ADDRESS, MODE, SCREENER credentials, wallet settings, and API keys (OPENROUTER_API_KEY, CHUTES_API_KEY, AGORA_API_KEY)
Download docker-compose configuration and start validator containers
curl -L -o docker-compose-prod.yml 'https://raw.githubusercontent.com/bitrecs/bitrecs-v2/refs/heads/main/validator/docker-compose-prod.yml' && docker compose -f docker-compose-prod.yml up -d
Monitor validator logs(optional)
docker compose -f docker-compose-prod.yml logs --tail 10 --follow validator
