typo

VerifiableRewardsForScalableLogicalReasoning.py

@@ -339,7 +339,7 @@ class VerifiableRewardsForScalableLogicalReasoning(evaluate.Metric):
             if not validation_program:
                 raise ValueError(f"Example {i} does not contain validation program field")
 
-            eval_inputs.append((prediction, validation_program, eval_configm, TIMEOUT))
+            eval_inputs.append((prediction, validation_program, eval_config, TIMEOUT))
 
         # if more than 1k predictions, we use multiprocessing to speed up the evaluation
         if len(eval_inputs) > 500:
@@ -349,7 +349,7 @@ class VerifiableRewardsForScalableLogicalReasoning(evaluate.Metric):
                 results = list(tqdm(
                     pool.starmap(_evaluate_with_prolog, eval_inputs),
                     total=len(eval_inputs),
-                    desc="Evaluating rules (parallel)"
+                    desc=f"Evaluating rules (parallel processing with {num_cpus} CPUs)"
                 ))
         else:
             # Evaluate in the main thread (no multiprocessing)