Docker Overview

🐳 JoeyLLM Docker Development Overview

JoeyLLM runs inside a Docker container to ensure consistency across all contributors and to enable scalable training on GPU clusters.

🧭 This setup has two parts — both are required to get started:

1️⃣ Install Docker on your system (plus GPU support if you have an NVIDIA GPU)
2️⃣ Pull and run the JoeyLLM container, which contains the full dev environment

---

1️⃣ Set Up Docker on Your System

Platform	GPU Support	Setup Guide
Linux	✅ Full	🐧 Linux Setup
Windows (WSL2)	✅ Partial	🪟 Windows Setup
macOS	❌ CPU-only	🍎 macOS Setup

🐧 Linux is strongly recommended for contributors working on training or internals.

---

2️⃣ Launch the JoeyLLM Development Container

Once Docker is installed and ready:

• 🚀 Launching the JoeyLLM Development Container

This will walk you through:

• Pulling the official image
• Creating and starting the container
• Cloning the repo and installing Python dependencies inside the container

---

🔍 Why Docker?

Using Docker ensures:

• ✅ Reproducible, isolated environments
• 🧩 No local dependency/version conflicts
• ⚙️ Compatibility with GPU clusters and inference pipelines
• 🤖 Easy GPU acceleration (where available)
• 🤝 Consistent dev workflow across all contributors

🛠️ Important:
JoeyLLM is built to run at scale across shared infrastructure.
Docker ensures your local environment matches our GPU cluster setup.

---

💡 Do I Need a GPU?

No — a GPU is not required.

• 🧪 Basic testing works fine on CPU.
• ☁️ Full training runs on shared GPU infrastructure.
• ⚡ If you have an NVIDIA GPU, enable GPU support for faster dev/testing (optional).

---

🧠 Questions?

Need help?

• 💬 Join our Discord
• 🐛 Open an Issue