new tasks

src/leaderboard/read_evals.py

@@ -106,7 +106,7 @@ class EvalResult:
             task = task.value
 
             task_n_shot_num = n_shot_num
-            if 'perplexity' in task.metric: # perplexity is the same for 0-shot and 5-shot and is calculated only with 0-shot
+            if 'perplexity' in task.metric or task.benchmark=='polish_eq_bench': # perplexity is the same for 0-shot and 5-shot and is calculated only with 0-shot
                 task_n_shot_num = 0
 
             # We average all scores of a given metric (not all metrics are present in all files)