add adamw to torch backend (#532)

Co-authored-by: Haifeng Jin <haifeng-jin@users.noreply.github.com>

keras_core/backend/torch/optimizers/torch_adamw.py

@@ -0,0 +1,6 @@
+from keras_core import optimizers
+from keras_core.backend.torch.optimizers import torch_adam
+
+
+class AdamW(torch_adam.Adam, optimizers.AdamW):
+    pass

keras_core/backend/torch/optimizers/torch_optimizer.py

@@ -8,11 +8,13 @@ class TorchOptimizer(BaseOptimizer):
     def __new__(cls, *args, **kwargs):
         # Import locally to avoid circular imports.
         from keras_core.backend.torch.optimizers import torch_adam
+        from keras_core.backend.torch.optimizers import torch_adamw
         from keras_core.backend.torch.optimizers import torch_sgd
 
         OPTIMIZERS = {
             optimizers.SGD: torch_sgd.SGD,
             optimizers.Adam: torch_adam.Adam,
+            optimizers.AdamW: torch_adamw.AdamW,
         }
         if cls in OPTIMIZERS:
             return OPTIMIZERS[cls](*args, **kwargs)

keras_core/optimizers/adamw.py

@@ -1,10 +1,10 @@
-from keras_core import ops
 from keras_core.api_export import keras_core_export
+from keras_core.optimizers import adam
 from keras_core.optimizers import optimizer
 
 
 @keras_core_export(["keras_core.optimizers.AdamW"])
-class AdamW(optimizer.Optimizer):
+class AdamW(adam.Adam):
     """Optimizer that implements the AdamW algorithm.
 
     AdamW optimization is a stochastic gradient descent method that is based on
@@ -68,99 +68,26 @@ def __init__(
     ):
         super().__init__(
             learning_rate=learning_rate,
+            beta_1=beta_1,
+            beta_2=beta_2,
+            epsilon=epsilon,
+            amsgrad=amsgrad,
             name=name,
+            weight_decay=weight_decay,
             clipnorm=clipnorm,
             clipvalue=clipvalue,
             global_clipnorm=global_clipnorm,
             use_ema=use_ema,
             ema_momentum=ema_momentum,
             ema_overwrite_frequency=ema_overwrite_frequency,
         )
-        self.weight_decay = weight_decay
-        self.beta_1 = beta_1
-        self.beta_2 = beta_2
-        self.epsilon = epsilon
-        self.amsgrad = amsgrad
 
         if self.weight_decay is None:
             raise ValueError(
                 "Argument `weight_decay` must be a float. Received: "
                 "weight_decay=None"
             )
 
-    def build(self, var_list):
-        """Initialize optimizer variables.
-
-        AdamW optimizer has 3 types of variables: momentums, velocities and
-        velocity_hat (only set when amsgrad is applied),
-
-        Args:
-            var_list: list of model variables to build AdamW variables on.
-        """
-        if self.built:
-            return
-        super().build(var_list)
-        self._momentums = []
-        self._velocities = []
-        for var in var_list:
-            self._momentums.append(
-                self.add_variable_from_reference(
-                    reference_variable=var, name="m"
-                )
-            )
-            self._velocities.append(
-                self.add_variable_from_reference(
-                    reference_variable=var, name="v"
-                )
-            )
-        if self.amsgrad:
-            self._velocity_hats = []
-            for var in var_list:
-                self._velocity_hats.append(
-                    self.add_variable_from_reference(
-                        reference_variable=var, name="vhat"
-                    )
-                )
-
-    def update_step(self, gradient, variable, learning_rate):
-        """Update step given gradient and the associated model variable."""
-        lr = ops.cast(learning_rate, variable.dtype)
-        gradient = ops.cast(gradient, variable.dtype)
-        local_step = ops.cast(self.iterations + 1, variable.dtype)
-        beta_1_power = ops.power(
-            ops.cast(self.beta_1, variable.dtype), local_step
-        )
-        beta_2_power = ops.power(
-            ops.cast(self.beta_2, variable.dtype), local_step
-        )
-
-        m = self._momentums[self._get_variable_index(variable)]
-        v = self._velocities[self._get_variable_index(variable)]
-
-        alpha = lr * ops.sqrt(1 - beta_2_power) / (1 - beta_1_power)
-
-        m.assign(m + (gradient - m) * (1 - self.beta_1))
-        v.assign(v + (ops.square(gradient) - v) * (1 - self.beta_2))
-        if self.amsgrad:
-            v_hat = self._velocity_hats[self._get_variable_index(variable)]
-            v_hat.assign(ops.maximum(v_hat, v))
-            v = v_hat
-        variable.assign(variable - (m * alpha) / (ops.sqrt(v) + self.epsilon))
-
-    def get_config(self):
-        config = super().get_config()
-
-        config.update(
-            {
-                "weight_decay": self.weight_decay,
-                "beta_1": self.beta_1,
-                "beta_2": self.beta_2,
-                "epsilon": self.epsilon,
-                "amsgrad": self.amsgrad,
-            }
-        )
-        return config
-
 
 AdamW.__doc__ = AdamW.__doc__.replace(
     "{{base_optimizer_keyword_args}}", optimizer.base_optimizer_keyword_args

keras_core/optimizers/adamw_test.py

@@ -4,6 +4,7 @@
 import numpy as np
 
 from keras_core import backend
+from keras_core import ops
 from keras_core import testing
 from keras_core.optimizers.adamw import AdamW
 
@@ -22,7 +23,7 @@ def test_config(self):
 
     def test_single_step(self):
         optimizer = AdamW(learning_rate=0.5)
-        grads = np.array([1.0, 6.0, 7.0, 2.0])
+        grads = ops.array([1.0, 6.0, 7.0, 2.0])
         vars = backend.Variable([1.0, 2.0, 3.0, 4.0])
         optimizer.apply_gradients(zip([grads], [vars]))
         self.assertAllClose(
@@ -31,7 +32,7 @@ def test_single_step(self):
 
     def test_weight_decay(self):
         grads, var1, var2, var3 = (
-            np.zeros(()),
+            ops.zeros(()),
             backend.Variable(2.0),
             backend.Variable(2.0, name="exclude"),
             backend.Variable(2.0),
@@ -55,8 +56,8 @@ def test_correctness_with_golden(self):
         optimizer = AdamW(learning_rate=1.0, weight_decay=0.5, epsilon=2)
 
         x = backend.Variable(np.ones([10]))
-        grads = np.arange(0.1, 1.1, 0.1)
-        first_grads = np.full((10,), 0.01)
+        grads = ops.arange(0.1, 1.1, 0.1)
+        first_grads = ops.full((10,), 0.01)
 
         # fmt: off
         golden = np.array(