Fix Sylvain's nits on the original KerasMetricCallback PR (#18300)

* Fix Sylvain's nits on the original PR

* Update src/transformers/keras_callbacks.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Re-add "optional" to docstring
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

src/transformers/keras_callbacks.py

@@ -84,8 +84,8 @@ class KerasMetricCallback(Callback):
         output_cols: Optional[List[str]] = None,
         label_cols: Optional[List[str]] = None,
         batch_size: Optional[int] = None,
-        predict_with_generate: Optional[bool] = False,
-        use_xla_generation: Optional[bool] = False,
+        predict_with_generate: bool = False,
+        use_xla_generation: bool = False,
         generate_kwargs: Optional[dict] = None,
     ):
         super().__init__()
@@ -138,7 +138,7 @@ class KerasMetricCallback(Callback):
             logging.warning("TF versions less than 2.7 may encounter issues with KerasMetricCallback!")
 
         self.use_xla_generation = use_xla_generation
-        self.generate_kwargs = generate_kwargs or {}
+        self.generate_kwargs = {} if generate_kwargs is None else generate_kwargs
 
         self.generation_function = None
 
@@ -202,11 +202,7 @@ class KerasMetricCallback(Callback):
             if self.use_xla_generation and self.generation_function is None:
 
                 def generation_function(inputs, attention_mask):
-                    return self.model.generate(
-                        inputs,
-                        attention_mask=attention_mask,
-                        **self.generate_kwargs,
-                    )
+                    return self.model.generate(inputs, attention_mask=attention_mask, **self.generate_kwargs)
 
                 self.generation_function = tf.function(generation_function, jit_compile=True)