⏳ Another player is currently using the stream. Please wait for them to finish.
💡 Tip: Click anywhere on this page to enable keyboard controls. Use ↑/↓ Arrow Keys or W/S to control the paddle.
This demo uses a small transformer model trained with rectified flow matching to simulate Pong game frames conditioned on user inputs. The model generates 24×24 pixel frames in real-time using diffusion sampling with configurable steps. Performance targets ~16 FPS with 4 diffusion steps on GPU hardware.