Backwards compat for decay arg

keras_core/optimizers/adadelta.py

@@ -50,6 +50,7 @@ class Adadelta(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="adadelta",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -61,6 +62,7 @@ class Adadelta(optimizer.Optimizer):
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
             name=name,
+            **kwargs,
         )
         self.rho = rho
         self.epsilon = epsilon

keras_core/optimizers/adafactor.py

@@ -57,6 +57,7 @@ class Adafactor(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="adafactor",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -68,6 +69,7 @@ class Adafactor(optimizer.Optimizer):
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
+            **kwargs,
         )
         self.beta_2_decay = beta_2_decay
         self.epsilon_1 = epsilon_1

keras_core/optimizers/adagrad.py

@@ -45,6 +45,7 @@ class Adagrad(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="adagrad",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -56,6 +57,7 @@ class Adagrad(optimizer.Optimizer):
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
             name=name,
+            **kwargs,
         )
         self.initial_accumulator_value = initial_accumulator_value
         self.epsilon = epsilon

keras_core/optimizers/adam.py

@@ -55,6 +55,7 @@ class Adam(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="adam",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -66,6 +67,7 @@ class Adam(optimizer.Optimizer):
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
+            **kwargs,
         )
         self.beta_1 = beta_1
         self.beta_2 = beta_2

keras_core/optimizers/adamax.py

@@ -64,6 +64,7 @@ class Adamax(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="adamax",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -75,6 +76,7 @@ class Adamax(optimizer.Optimizer):
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
+            **kwargs,
         )
         self.beta_1 = beta_1
         self.beta_2 = beta_2

keras_core/optimizers/adamw.py

@@ -65,6 +65,7 @@ class AdamW(adam.Adam):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="adamw",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -80,6 +81,7 @@ class AdamW(adam.Adam):
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
+            **kwargs,
         )
 
         if self.weight_decay is None:

keras_core/optimizers/base_optimizer.py

@@ -22,9 +22,17 @@ class BaseOptimizer:
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name=None,
+        **kwargs,
     ):
         self._lock = False
 
+        if kwargs.pop("decay", None) is not None:
+            warnings.warn(
+                "Argument `decay` is no longer supported and will be ignored."
+            )
+        if kwargs:
+            raise ValueError(f"Argument(s) not recognized: {kwargs}")
+
         self.name = name
         self.weight_decay = weight_decay
         self.clipnorm = clipnorm

keras_core/optimizers/ftrl.py

@@ -92,6 +92,7 @@ class Ftrl(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="ftrl",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -103,6 +104,7 @@ class Ftrl(optimizer.Optimizer):
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
+            **kwargs,
         )
 
         if initial_accumulator_value < 0.0:

keras_core/optimizers/lion.py

@@ -54,6 +54,7 @@ class Lion(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="lion",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -65,6 +66,7 @@ class Lion(optimizer.Optimizer):
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
+            **kwargs,
         )
         self.beta_1 = beta_1
         self.beta_2 = beta_2

keras_core/optimizers/nadam.py

@@ -50,6 +50,7 @@ class Nadam(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="nadam",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -61,6 +62,7 @@ class Nadam(optimizer.Optimizer):
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
+            **kwargs,
         )
         self.beta_1 = beta_1
         self.beta_2 = beta_2

keras_core/optimizers/rmsprop.py

@@ -65,6 +65,7 @@ class RMSprop(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=100,
         name="rmsprop",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -76,6 +77,7 @@ class RMSprop(optimizer.Optimizer):
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
             name=name,
+            **kwargs,
         )
         self.rho = rho
         self.momentum = momentum

keras_core/optimizers/sgd.py

@@ -53,6 +53,7 @@ class SGD(optimizer.Optimizer):
         ema_momentum=0.99,
         ema_overwrite_frequency=None,
         name="SGD",
+        **kwargs,
     ):
         super().__init__(
             learning_rate=learning_rate,
@@ -64,6 +65,7 @@ class SGD(optimizer.Optimizer):
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
+            **kwargs,
         )
         if not isinstance(momentum, float) or momentum < 0 or momentum > 1:
             raise ValueError("`momentum` must be a float between [0, 1].")