diff --git a/evaluation/data_processing/answer_extraction.py b/evaluation/data_processing/answer_extraction.py
index c2691d9..14c0b23 100644
--- a/evaluation/data_processing/answer_extraction.py
+++ b/evaluation/data_processing/answer_extraction.py
@@ -127,7 +127,7 @@ def strip_string(string):
     string = string.replace("infinity", "\\infty")
     if "\\infty" not in string:
         string = string.replace("inf", "\\infty")
-    string = string.replace("+\\inity", "\\infty")
+    string = string.replace("+\\infty", "\\infty")
 
     # and 
     # string = string.replace("and", "")
@@ -305,7 +305,6 @@ def extract_ocwcourses_few_shot_answer(question, reasoning, task):
     patt = regex.search(r"final answer is (?P<ans>.*)\. I hope it is correct.", reasoning)
     if patt is None:
         pred = "[invalid]"
-        print(f"DEBUG >>>\n{reasoning}", flush=True)
     else:
         pred = patt.group('ans')
     return pred
@@ -331,7 +330,6 @@ def extract_cmath_few_shot_test(question, reasoning, task):
         try:
             ans = [s for s in regex.findall(r'-?\d+\.?\d*', ans)][-1]
         except:
-            print(f"DEBUG CMATH: {reasoning}", flush=True)
             ans = "[invalid]"
     else:
         ans = extract_last_single_answer(question, reasoning, task)
diff --git a/evaluation/eval/eval_script.py b/evaluation/eval/eval_script.py
index 0718130..3ab13fd 100644
--- a/evaluation/eval/eval_script.py
+++ b/evaluation/eval/eval_script.py
@@ -19,9 +19,6 @@ def is_correct(item, pred_key='prediction', prec=1e-3):
                 if is_correct(item_cpy, pred_key=pred_key, prec=prec):
                     pred_matched.add(i)
                     ans_matched.add(j)
-                    if item_cpy[pred_key] == '2,3,4':
-                        print(item, flush=True)
-                        print("wtf", flush=True)
         return len(pred_matched) == len(pred) and len(ans_matched) == len(ans)
     elif isinstance(pred, str) and isinstance(ans, str):
         if '\\cup' in pred and '\\cup' in ans:
@@ -40,8 +37,7 @@ def is_correct(item, pred_key='prediction', prec=1e-3):
             label = label or (ans and pred == ans) or math_equal(pred, ans)
             return label
     else:
-        print(item, flush=True)
-        raise NotImplementedError()
+        raise NotImplementedError(f"Unsupported types: pred={type(pred)}, ans={type(ans)}")
 
 def eval_math(item, pred_key='prediction', prec=1e-3):
     pred = item[pred_key]