Quick Start

Get your NeuroShard node running and earning NEURO in under 5 minutes.

Prerequisites

• Python 3.10+ installed
• 4GB+ RAM (more RAM = more layers = more rewards)
• Internet connection for peer discovery and training sync

Installation

Install via pip

pip install nexaroa

For NVIDIA GPU support, install PyTorch with CUDA first:

pip install torch --index-url https://download.pytorch.org/whl/cu118
pip install nexaroa

Get Your Wallet Token

1. Go to neuroshard.com/signup
2. Create an account
3. Navigate to your Dashboard → Wallet
4. Copy your wallet token (64-character hex string or 12-word mnemonic)

Keep Your Token Safe

Your wallet token is the key to your NEURO earnings. Never share it with anyone. Store it securely.

Start Your Node

neuroshard --token YOUR_WALLET_TOKEN

That's it! Your node will:

1. Detect available memory and GPU automatically
2. Register with the network and get layer assignments
3. Start training NeuroLLM and earning NEURO
4. Open a dashboard at http://localhost:8000/

Verify It's Working

Check the Dashboard

Open http://localhost:8000/ in your browser. You should see:

• Your node ID and network status
• Layers assigned to your node
• Training progress and loss
• NEURO balance and earnings rate

Check the Logs

# Look for these messages:
[NODE] ✅ Wallet recovered from mnemonic
[NODE] Starting on port 8000...
[NODE] Dashboard: http://localhost:8000/
[NODE] Assigned 24 layers: [0, 1, 2, ...]
[NODE] GPU detected: CUDA (NVIDIA GeForce RTX 3080)
[GENESIS] Data loader ready: 1024 shards available

Check the Ledger

Visit neuroshard.com/ledger and search for your wallet address to see your balance and transaction history.

Common Options

# Use a custom port
neuroshard --token YOUR_TOKEN --port 9000

# Limit memory usage (in MB)
neuroshard --token YOUR_TOKEN --memory 4096

# Inference-only mode (no training)
neuroshard --token YOUR_TOKEN --no-training

# Run without opening browser
neuroshard --token YOUR_TOKEN --no-browser

# Set CPU thread limit
neuroshard --token YOUR_TOKEN --cpu-threads 4

Expected Earnings

Your earnings depend on:

• Memory: More RAM = more layers = higher rewards
• Training Activity: Active training earns ~300x more than idle
• Role: Drivers (Layer 0) and Validators (Last Layer) earn bonuses

Node Type	Memory	Daily Earnings (Active)
Raspberry Pi	2GB	~10-20 NEURO
Laptop	8GB	~40-60 NEURO
Gaming PC	16GB	~80-120 NEURO
Server	64GB+	~200-400 NEURO

Earnings Fluctuate

Earnings depend on network activity, model quality, and inference demand. These are estimates based on typical conditions.

Troubleshooting

"No GPU detected"

Install CUDA-enabled PyTorch:

pip uninstall torch
pip install torch --index-url https://download.pytorch.org/whl/cu121

"Insufficient memory"

Limit the memory usage:

neuroshard --token YOUR_TOKEN --memory 2048

"Connection refused"

Check your firewall settings. NeuroShard uses:

• Port 8000 (HTTP dashboard and REST API)
• Port 9000 (gRPC for peer communication)

The gRPC port is always HTTP port + 1000. If you change the HTTP port with --port, the gRPC port changes accordingly.

"Data not ready"

The Genesis data loader needs time to download training shards. Wait 30-60 seconds for initialization.

Next Steps

• Running a Node — Detailed node configuration
• Network Roles — Understand Driver, Worker, Validator roles
• CLI Reference — All command-line options
• NEURO Economics — Maximize your earnings