fix single eval

VerifiableRewardsForScalableLogicalReasoning.py

@@ -246,6 +246,7 @@ def extract_ilp_from_text_v2(text, target_predicate=None, allow_multiple_rules=F
     # Pre-process: collapse code blocks to single lines
     text = re.sub(r'\n\s*', ' ', text)  # crude: flatten all to one line
     # Rule pattern, across newlines
+    # rule_pattern = re.compile(rf'({target_predicate}\([^()]*\)\s*:-.*?\.)')
     rule_pattern = re.compile(rf'({target_predicate}\([^()]*\)\s*:-.*?\.)')
     rules = list(rule_pattern.findall(text))
     if len(rules) > 1 and not allow_multiple_rules:
@@ -353,8 +354,8 @@ class VerifiableRewardsForScalableLogicalReasoning(evaluate.Metric):
         else:
             # Evaluate in the main thread (no multiprocessing)
             results = []
-            for prediction, validation_program, eval_config in tqdm(eval_inputs, total=len(predictions), desc="Evaluating rules"):
-                results.append(_evaluate_with_prolog(prediction, validation_program, eval_config))
+            for prediction, validation_program, eval_config, t in tqdm(eval_inputs, total=len(predictions), desc="Evaluating rules"):
+                results.append(_evaluate_with_prolog(prediction, validation_program, eval_config, timeout=t))
 
         # Calculate metrics
         partial_scores = [result["partial_score"] for result in results]