update a bit one

app.py

@@ -2211,8 +2211,16 @@ with block:
     - **Number of Models**: {NUM_MODELS}
     - **Mode of Evaluation**: Zero-Shot, Five-Shot
     
+    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
+    Know Issues:
+    - For base models, the output of base model is not truncated as no EOS detected. Evaluation could be affected, especially with length-aware metrics.
+
+    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
     The following table shows the performance of the models on the SeaEval benchmark.
+    - For **Zero-shot** performance, it is the median value from 5 distinct prompts shown on the above leaderboard to mitigate the influence of random variations induced by prompts.
+    - (-1) value indicates the results are ready yet.
     - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
+
     """)
 
 
@@ -3221,13 +3229,7 @@ with block:
 
 
     gr.Markdown(r"""
-
-    - For **Zero-shot** performance, it is the median value from 5 distinct prompts shown on the above leaderboard to mitigate the influence of random variations induced by prompts.
-
-    - (-1) value indicates the results are ready yet and will be updated soon.
-
-
-    
+   
     If our datasets and leaderboard are useful, please consider cite:
     ```bibtex
         @article{SeaEval,