Gensyn Testnet Node Setup Guide

🎯 Overview

This comprehensive guide helps you set up and manage a Gensyn testnet node on a GPU server. The Gensyn network is a decentralized compute network for AI training, and running a node allows you to participate in the network and earn rewards.

💻 System Requirements

Component	Minimum Requirement	Recommended
CPU	arm64 or amd64	amd64
RAM	25 GB	32 GB+
GPU	RTX 3090, RTX 4090, A100, or H100	RTX 4090 or A100
Storage	50 GB	100 GB+
Python	3.10 or higher	3.11+
Network	Stable internet connection	High-speed connection

🚀 Quick Setup

Step 1: Rent a GPU Server

1. Visit Quick Pod: Go to Quick Pod Console
2. Create Account: Sign up with your email and verify your account
3. Add Funds: Deposit funds using crypto or credit card
4. Select Template: Go to Templates → Select CUDA 12.6
5. Configure Docker: Clone the template and edit Docker options:

   -p 8888:8888 -p 3000:3000

6. Choose GPU: Select RTX 4090 (recommended) or RTX 3090
7. Create POD: Click Create POD and wait for initialization

Step 2: Connect and Prepare Server

1. Access Terminal: Click Connect → Web Terminal

2. Install System Dependencies:

   # Update package lists
   apt update && apt install -y sudo
   sudo apt update
   
   # Install essential packages
   sudo apt install -y python3 python3-venv python3-pip curl wget screen git lsof htop
   
   # Install Yarn for Node.js dependencies
   curl -sS https://dl.yarnpkg.com/debian/pubkey.gpg | sudo apt-key add -
   echo "deb https://dl.yarnpkg.com/debian/ stable main" | sudo tee /etc/apt/sources.list.d/yarn.list
   sudo apt update && sudo apt install -y yarn

3. Install Node.js:

   curl -sSL https://raw.githubusercontent.com/zunxbt/installation/main/node.sh | bash

Step 3: Install and Configure Gensyn Node

1. Clone Repository:

   cd $HOME
   [ -d rl-swarm ] && rm -rf rl-swarm
   git clone https://github.com/gensyn-ai/rl-swarm.git
   cd rl-swarm

2. Create Virtual Environment:

   python3 -m venv .venv
   source .venv/bin/activate

3. Start tmux Session:

   tmux new-session -d -s gensyn
   tmux attach-session -t gensyn

4. Run the Swarm:

   ./run_rl_swarm.sh

5. Complete Setup Prompts:

   • Connect to Testnet? → Type Y and press Enter
   • Push models to Hugging Face? → Type Y and press Enter
   • Hugging Face token → Paste your write token (get from Hugging Face Settings)

6. Detach from tmux: Press Ctrl + B, then D

🔧 Advanced Setup

Using the Auto-Restart Script

For better reliability, use the included auto-restart script:

1. Make Script Executable:

   chmod +x script.sh

2. Start Auto-Restart Monitoring:

   ./script.sh

3. Useful Commands:

   ./script.sh status    # Check node status
   ./script.sh logs      # View logs
   ./script.sh clean     # Clean up processes
   ./script.sh stop      # Stop auto-restart

Managing tmux Sessions

# List sessions
tmux list-sessions

# Attach to session
tmux attach-session -t gensyn

# Detach from session (while inside)
Ctrl + B, then D

# Kill session
tmux kill-session -t gensyn

🔐 Security and Key Management

⚠️ Critical Warnings

• NEVER lose your swarm.pem file - it's your node's unique identity
• Always use the same email for your node to maintain consistency
• Keep backups of your swarm.pem in multiple secure locations
• Never share your private keys with anyone

Managing Your Swarm Key

Method 1: Jupyter Lab (Recommended)

1. Access Jupyter Lab: Open http://your-server-ip:8888 in your browser
2. Upload/Download: Use the file browser to manage your swarm.pem
3. Backup: Download and store in secure location

Method 2: Direct File Transfer

# Download from server
scp username@server-ip:~/rl-swarm/swarm.pem ./swarm.pem

# Upload to server
scp ./swarm.pem username@server-ip:~/rl-swarm/swarm.pem

Method 3: Cloud Storage

# Upload to cloud storage (example with rclone)
rclone copy swarm.pem remote:backups/

# Download from cloud storage
rclone copy remote:backups/swarm.pem ./

🌐 Remote Access Setup

Option 1: Cloudflare Tunnel (Recommended)

1. Install Cloudflare Tunnel:

   wget https://github.com/cloudflare/cloudflared/releases/latest/download/cloudflared-linux-amd64.deb
   sudo dpkg -i cloudflared-linux-amd64.deb

2. Create Tunnel:

   cloudflared tunnel --url http://localhost:3000

3. Follow Authentication: Complete the Cloudflare authentication process

4. Access: Use the provided public URL to access your node

Option 2: SSH Port Forwarding

# Forward local port to remote server
ssh -L 3000:localhost:3000 username@server-ip

# Then access via http://localhost:3000

📊 Monitoring Your Node

Check Node Status

# Using the script
./script.sh status

# Manual checks
tmux list-sessions
pgrep -af "python.*run_rl_swarm"
nvidia-smi  # Check GPU usage

View Logs

# Real-time logs
./script.sh logs

# Manual log viewing
tail -f ~/gensyn_auto_restart.log

Monitor Resources

# GPU usage
nvidia-smi -l 1

# System resources
htop

# Disk usage
df -h

🔧 Troubleshooting

Common Issues

Node Won't Start

# Check if tmux session exists
tmux list-sessions

# Kill existing sessions
tmux kill-session -t gensyn

# Clean up processes
pkill -f "python.*run_rl_swarm"
pkill -f "python.*main"

# Restart
./script.sh start

Authentication Issues

# Remove cached auth data
rm -rf modal-login/temp-data/*.json

# Re-authenticate
./run_rl_swarm.sh

GPU Memory Issues

# Check GPU memory
nvidia-smi

# Restart if memory is stuck
sudo systemctl restart nvidia-persistenced

Network Issues

# Check connectivity
ping 8.8.8.8

# Check DNS
nslookup google.com

# Restart network (if needed)
sudo systemctl restart networking

Performance Optimization

1. GPU Memory: Ensure you have at least 2GB free VRAM
2. CPU Priority: Run with nice priority for better performance
3. Network: Use stable, high-speed internet connection
4. Storage: Use SSD storage for better I/O performance

📈 Earning and Rewards

How Rewards Work

• Participation: Earn rewards by participating in training rounds
• Performance: Better hardware = higher potential rewards
• Uptime: More uptime = more opportunities to earn
• Network: Rewards distributed based on contribution to the network

Tracking Your Earnings

• Monitor your node's participation in the Gensyn dashboard
• Check logs for successful round completions
• Track your Hugging Face model uploads

🆘 Support and Community

Getting Help

• Documentation: Gensyn Docs
• Discord: Gensyn Community
• Twitter: @0xemir_
• GitHub: rl-swarm Repository

Useful Commands Reference

# Node management
./script.sh start      # Start node
./script.sh stop       # Stop auto-restart
./script.sh status     # Check status
./script.sh logs       # View logs
./script.sh clean      # Clean processes

# tmux management
tmux list-sessions     # List sessions
tmux attach -t gensyn  # Attach to session
tmux kill-session -t gensyn  # Kill session

# System monitoring
nvidia-smi             # GPU status
htop                   # System resources
df -h                  # Disk usage
free -h                # Memory usage

📺 Video Tutorial

Gensyn.Tutorial.mp4

---

Happy mining! 🚀

Remember to star the repository: https://github.com/gensyn-ai/rl-swarm