🐛 Troubleshooting

Service won't start	Prediction fails with timeout	Model loading fails
Check Docker logs: docker-compose logs	Increase RL_INFERENCE_SERVICE_TIMEOUT	Verify model path in environment variables
Verify model file exists and is correct format	Check if Python service is running: docker-compose ps	Ensure model.zip is a valid stable-baselines3 PPO model
Check port availability (8000, 8080)	Verify network connectivity: docker-compose exec java-gateway ping rl-inference	Check file permissions

Also see Support