add thinks

about.py

@@ -41,5 +41,6 @@ METRIC_INFO_TEXT = """
 
 - **Meter** - How closely the model sticks to the meter of the lines.
 - **Verse** - How closely the model aligns the lines to the verse and chorus breakup.
-- **Focus** - How much of the response is extraneous commentary instead of the song. Focus in particular has a very minor contribution to the final score
+- **Focus** - How much of the response is extraneous commentary instead of the song. *(Focus in particular has a very minor contribution to the final score)*
+- **Thinking** - The **averaged** number of thinking tokens per response. Zero means it's not a reasoning model.
 """

utils.py

@@ -68,6 +68,7 @@ def load_leaderboard() -> pd.DataFrame:
             "meter_score": "Meter",
             "verse_score": "Verse",
             "focus_score": "Focus",
+            "avg_think_tokens": "Thinking",
             "evaluated_at": "Evaluated At",
         }
         display_cols = [c for c in col_map if c in df.columns]
@@ -116,12 +117,9 @@ def load_leaderboard() -> pd.DataFrame:
                 "Meter": "{:.1%}",
                 "Verse": "{:.1%}",
                 "Focus": "{:.1%}",
+                "Thinking": "{:0.0f}",
             })
             .hide(axis="index")
-            .set_table_styles([
-                dict(selector='th', props=[('text-align', 'center')]),
-                dict(selector='td', props=[('text-align', 'center')]),
-            ])
         )        
 
         return styled
@@ -140,6 +138,7 @@ def empty_leaderboard() -> pd.DataFrame:
             "Meter",
             "Verse",
             "Focus",
+            "Thinking",
             "Evaluated At",
         ]
     )