Fix metric.run_metric out of the try-except-else

utilities.py

@@ -185,13 +185,18 @@ def evaluate_answer(df):
             target = row["target_answer"]
             predicted = row["predicted_answer"]
             try:
-                result = metric.run_metric(target = target, prediction = predicted)
+                predicted = eval(predicted)
             except Exception as e:
-                result = None
+                result = 0
+            else:
+                result = metric.run_metric(target = target, prediction = predicted)
             results.append(result)
         df[metric_name] = results
     return df
 
+
+
+
 models = [
           "gpt-4o-mini",
           "deepseek-ai/DeepSeek-R1-Distill-Llama-70B",