Fix the rollout clock ordering

README.md

@@ -92,8 +92,9 @@ for step in range(20):  # 5 days
     with torch.no_grad():
         outputs = model(**inputs)
     forecast = processor.postprocess(outputs.prediction, state)
-    valid_time = valid_time + step_seconds
+    # `valid_time` is the time this forecast is valid at, so it stamps the appended frame first.
     state = processor.advance_state(state, forecast, valid_time)
+    valid_time = valid_time + step_seconds
 ```
 
 Turning the gridded cyclone diagnostics into tracks requires the tracker from the