Support transformers v4.49 logits_to_keep for IPEX (huggingface#1188)

* fix logits_to_keep

Signed-off-by: jiqing-feng <jiqing.feng@intel.com>

* fix typo

Signed-off-by: jiqing-feng <jiqing.feng@intel.com>

* Update optimum/intel/ipex/modeling_base.py

Co-authored-by: Ella Charlaix <80481427+echarlaix@users.noreply.github.com>

* fix

Signed-off-by: jiqing-feng <jiqing.feng@intel.com>

* add comments

Signed-off-by: jiqing-feng <jiqing.feng@intel.com>

* fix format

Signed-off-by: jiqing-feng <jiqing.feng@intel.com>

---------

Signed-off-by: jiqing-feng <jiqing.feng@intel.com>
Co-authored-by: Ella Charlaix <80481427+echarlaix@users.noreply.github.com>

Author: jiqing-feng

optimum/intel/ipex/modeling_base.py

@@ -353,8 +353,17 @@ def _reorder_cache(self, *args, **kwargs):
     def prepare_inputs_for_generation(self, *args, **kwargs):
         return self.model.prepare_inputs_for_generation(*args, **kwargs)
 
+    def _supports_logits_to_keep(self) -> bool:
+        """
+        Return True if the current model supports the keyword argument `logits_to_keep` in forward()
+        to save memory. Checking it in this way allows to avoid using a new model attribute.
+        """
+        return "logits_to_keep" in set(inspect.signature(self.model.forward).parameters.keys())
+
     def _supports_num_logits_to_keep(self) -> bool:
         """
+        Will be deprecated after we no longer support transformers < 4.49
+
         Return True if the current model supports the keyword argument `num_logits_to_keep` in forward()
         to save memory. Checking it in this way allows to avoid using a new model attribute.
         """
@@ -470,8 +479,17 @@ def prepare_inputs_for_generation(self, *args, **kwargs):
     def get_encoder(self, *args, **kwargs):
         return self.model.get_encoder(*args, **kwargs)
 
+    def _supports_logits_to_keep(self) -> bool:
+        """
+        Return True if the current model supports the keyword argument `logits_to_keep` in forward()
+        to save memory. Checking it in this way allows to avoid using a new model attribute.
+        """
+        return "logits_to_keep" in set(inspect.signature(self.model.forward).parameters.keys())
+
     def _supports_num_logits_to_keep(self) -> bool:
         """
+        Will be deprecated after we no longer support transformers < 4.49
+
         Return True if the current model supports the keyword argument `num_logits_to_keep` in forward()
         to save memory. Checking it in this way allows to avoid using a new model attribute.
         """