🧠 OpenGym Copilot

OpenGym Copilot is a real-time visualizer and debug tool for OpenAI Gym environments. It lets you stream agent rollouts live from a backend, visualize frame-by-frame behavior, and inspect episodic outcomes — all through a clean, interactive frontend.

🚀 Reinforcement learning visualization shouldn't be guesswork.

---

🎯 Features

• 🔁 Real-time environment rollout streaming over WebSocket
• 🖼 Episode frame playback with looped simulation in-browser
• 📦 Environment selector (e.g., CartPole-v1, MountainCar-v0)
• 🧠 Built for easy Copilot & training integration (coming soon)
• 🟰 Parallelizable environments!!!
• 🎛 Adjustable frame capture interval (e.g., every N episodes)

---

🖥 Demo Preview

---

🛠 Tech Stack

• Frontend: React + Vite + WebSocket
• Backend: FastAPI + Gymnasium
• Visualization: JPEG frame streaming + canvas rendering (future: Three.js)
• Agent: Random policy for now (training coming soon)

---

🚀 Getting Started

🔧 Backend (Python + FastAPI)

pip install fastapi uvicorn gymnasium opencv-python
uvicorn main:app --reload

🔧 Frontend (npm + ReACT)

npm install
npm run dev