Improve technical description text

README.md

@@ -31,7 +31,7 @@ A real-time Pong game where frames are generated by a diffusion model trained wi
 
 ## Technical Details
 
-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.
+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 ~12 FPS with 4 diffusion steps on GPU hardware.
 
 ## Model Architecture
 

static/index.html

@@ -59,8 +59,8 @@
       <div id="waitingMessage" style="margin-top: 12px; padding: 8px; background: #333; border-radius: 4px; display: none; color: #ffa500;">
         ⏳ Another player is currently using the stream. Please wait for them to finish.
       </div>
-      <div>
-        This is the output of a small frame-autoregressive transformer trained with rectified flow matching to simulate pong frames conditioned on user inputs for the blue paddle. It should reach 12 FPS when using 4 steps for generation unless something else is running on my machine.
+      <div style="margin-top: 12px; font-size: 14px; color: #aaa; line-height: 1.5;">
+        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 ~12 FPS with 4 diffusion steps on GPU hardware.
       </div>
     </div>