add "_get_attr_from_logit_processors"

model.py

@@ -1846,7 +1846,14 @@ def _pad_to_max_length(
 
     sequences = torch.stack(sequences, dim=0)
     return sequences  
-
+    
+def _get_attr_from_logit_processors(logits_processor, logit_processor_class, attribute_name):
+    if logits_processor is not None:
+        logit_processor = next((cls for cls in logits_processor if isinstance(cls, logit_processor_class)), None)
+        if logit_processor:
+            return getattr(logit_processor, attribute_name, None)
+    return None
+    
 # CUSTOM (patch the generation method)    
 class CustomWhisperGenerationMixin(WhisperGenerationMixin):
     def generate(