PaddlePaddle
diff --git a/‎python/paddle/amp/grad_scaler.py
Copy file name to clipboard
+6-6Lines changed: 6 additions & 6 deletions b/‎python/paddle/amp/grad_scaler.py
Copy file name to clipboard
+6-6Lines changed: 6 additions & 6 deletions
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/dgc_optimizer.py
Copy file name to clipboard
+2-1Lines changed: 2 additions & 1 deletion b/‎python/paddle/distributed/fleet/meta_optimizers/dgc_optimizer.py
Copy file name to clipboard
+2-1Lines changed: 2 additions & 1 deletion
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/fp16_allreduce_optimizer.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion b/‎python/paddle/distributed/fleet/meta_optimizers/fp16_allreduce_optimizer.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/gradient_merge_optimizer.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion b/‎python/paddle/distributed/fleet/meta_optimizers/gradient_merge_optimizer.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/lamb_optimizer.py
Copy file name to clipboard
+3-3Lines changed: 3 additions & 3 deletions b/‎python/paddle/distributed/fleet/meta_optimizers/lamb_optimizer.py
Copy file name to clipboard
+3-3Lines changed: 3 additions & 3 deletions
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/lars_optimizer.py
Copy file name to clipboard
+3-2Lines changed: 3 additions & 2 deletions b/‎python/paddle/distributed/fleet/meta_optimizers/lars_optimizer.py
Copy file name to clipboard
+3-2Lines changed: 3 additions & 2 deletions
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/localsgd_optimizer.py
Copy file name to clipboard
-4Lines changed: 0 additions & 4 deletions b/‎python/paddle/distributed/fleet/meta_optimizers/localsgd_optimizer.py
Copy file name to clipboard
-4Lines changed: 0 additions & 4 deletions
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/meta_optimizer_base.py
Copy file name to clipboard
+2-2Lines changed: 2 additions & 2 deletions b/‎python/paddle/distributed/fleet/meta_optimizers/meta_optimizer_base.py
Copy file name to clipboard
+2-2Lines changed: 2 additions & 2 deletions
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/parameter_server_optimizer.py
Copy file name to clipboard
+1-2Lines changed: 1 addition & 2 deletions b/‎python/paddle/distributed/fleet/meta_optimizers/parameter_server_optimizer.py
Copy file name to clipboard
+1-2Lines changed: 1 addition & 2 deletions
diff --git a/‎python/paddle/distributed/fleet/meta_optimizers/pipeline_optimizer.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion b/‎python/paddle/distributed/fleet/meta_optimizers/pipeline_optimizer.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion
@@ -75,8 +75,8 @@ class AmpScaler:
 
         data = np.random.uniform(-1, 1, [10, 3, 32, 32]).astype('float32')
         model = paddle.nn.Conv2D(3, 2, 3)
-        optimizer = paddle.optimizer.SGDOptimizer(
-                learning_rate=0.01, parameter_list=model.parameters())
+        optimizer = paddle.optimizer.SGD(
+                learning_rate=0.01, parameters=model.parameters())
         scaler = paddle.amp.AmpScaler(init_loss_scaling=1024)
         data = paddle.to_tensor(data)
         with paddle.amp.amp_guard():
@@ -168,8 +168,8 @@ def scale(self, var):
 
                 data = np.random.uniform(-1, 1, [10, 3, 32, 32]).astype('float32')
                 model = paddle.nn.Conv2D(3, 2, 3)
-                optimizer = paddle.optimizer.SGDOptimizer(
-                        learning_rate=0.01, parameter_list=model.parameters())
+                optimizer = paddle.optimizer.SGD(
+                        learning_rate=0.01, parameters=model.parameters())
                 scaler = paddle.amp.AmpScaler(init_loss_scaling=1024)
                 data = paddle.to_tensor(data)
                 with paddle.amp.amp_guard():
@@ -221,8 +221,8 @@ def minimize(self, optimizer, *args, **kwargs):
 
                 data = np.random.uniform(-1, 1, [10, 3, 32, 32]).astype('float32')
                 model = paddle.nn.Conv2D(3, 2, 3)
-                optimizer = paddle.optimizer.SGDOptimizer(
-                        learning_rate=0.01, parameter_list=model.parameters())
+                optimizer = paddle.optimizer.SGD(
+                        learning_rate=0.01, parameters=model.parameters())
                 scaler = paddle.amp.AmpScaler(init_loss_scaling=1024)
                 data = paddle.to_tensor(data)
                 with paddle.amp.amp_guard():
 
@@ -21,9 +21,10 @@
 import paddle
 from paddle.common_ops_import import LayerHelper
 from paddle.fluid.dygraph import base as imperative_base
-from paddle.fluid.optimizer import Momentum, Optimizer
+from paddle.fluid.optimizer import Optimizer
 from paddle.framework import core, in_dynamic_mode
 from paddle.nn.clip import ClipGradByNorm, append_gradient_clip_ops
+from paddle.optimizer import Momentum
 from paddle.regularizer import L1Decay, L2Decay
 from paddle.static import create_global_var
 
 
@@ -152,6 +152,6 @@ def fp16_compression(param_and_grads):
 
     def apply_optimize(self, loss, startup_program, params_grads):
         new_params_grads = self.fp16_compression(params_grads)
-        return self.inner_opt.apply_optimize(
+        return self.inner_opt._apply_optimize(
             loss, startup_program=startup_program, params_grads=new_params_grads
         )
@@ -11,7 +11,7 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 
-from paddle.fluid.optimizer import GradientMergeOptimizer as GM
+from paddle.incubate.optimizer import GradientMergeOptimizer as GM
 
 from .meta_optimizer_base import MetaOptimizerBase
 
 
@@ -14,7 +14,7 @@
 import logging
 
 import paddle
-from paddle.fluid.optimizer import AdamOptimizer
+from paddle.optimizer import Adam
 
 from .meta_optimizer_base import MetaOptimizerBase
 
@@ -38,7 +38,7 @@ def _set_basic_info(
         )
 
         opt = self.inner_opt
-        if not isinstance(opt, AdamOptimizer):
+        if not isinstance(opt, Adam):
             return
 
         configs = self.user_defined_strategy.lamb_configs
@@ -72,7 +72,7 @@ def _can_apply(self):
             return False
 
         if self.user_defined_strategy.lamb:
-            if not isinstance(self.inner_opt, AdamOptimizer):
+            if not isinstance(self.inner_opt, Adam):
                 logging.warn(
                     "lamb need the inner optimizer to be AdamOptimizer optimizer but got {}.".format(
                         self.inner_opt.type
 
@@ -13,7 +13,8 @@
 
 import logging
 
-from paddle.fluid.optimizer import LarsMomentumOptimizer, Momentum
+from paddle.incubate.optimizer import LarsMomentumOptimizer
+from paddle.optimizer import Momentum
 
 from .meta_optimizer_base import MetaOptimizerBase
 
@@ -98,7 +99,7 @@ def apply_gradients(self, params_grads):
         return self.lars_opt.apply_gradients(params_grads=params_grads)
 
     def apply_optimize(self, loss, startup_program, params_grads):
-        return self.lars_opt.apply_optimize(
+        return self.lars_opt._apply_optimize(
             loss, startup_program=startup_program, params_grads=params_grads
         )
 
 
@@ -49,9 +49,7 @@ def _can_apply(self):
             self.inner_opt,
             (
                 paddle.optimizer.momentum.Momentum,
-                paddle.fluid.optimizer.Momentum,
                 paddle.optimizer.sgd.SGD,
-                paddle.fluid.optimizer.SGD,
             ),
         )
 
@@ -235,9 +233,7 @@ def _can_apply(self):
             self.inner_opt,
             (
                 paddle.optimizer.Momentum,
-                paddle.fluid.optimizer.Momentum,
                 paddle.optimizer.sgd.SGD,
-                paddle.fluid.optimizer.SGD,
             ),
         )
 
 
@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from paddle.fluid.optimizer import Optimizer
+from paddle.optimizer import Optimizer
 
 __all__ = []
 
@@ -81,7 +81,7 @@ def backward(
         )
 
     def apply_optimize(self, loss, startup_program, params_grads):
-        return self.inner_opt.apply_optimize(
+        return self.inner_opt._apply_optimize(
             loss, startup_program=startup_program, params_grads=params_grads
         )
 
 
@@ -17,7 +17,6 @@
 import subprocess
 
 import paddle
-from paddle import fluid
 from paddle.framework import core
 
 from ..base.private_helper_function import wait_server_ready
@@ -293,7 +292,7 @@ def get_sys_free_mem():
                     % (platform.system())
                 )
 
-        if not isinstance(self.inner_opt, fluid.optimizer.SGDOptimizer):
+        if not isinstance(self.inner_opt, paddle.optimizer.SGD):
             return False
 
         free = get_sys_free_mem()
 
@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 
 import paddle
-from paddle.fluid.optimizer import PipelineOptimizer as PO
+from paddle.incubate.optimizer import PipelineOptimizer as PO
 
 from .common import (
     OP_ROLE_KEY,
Original file line number	Diff line number	Diff line change
`@@ -152,6 +152,6 @@ def fp16_compression(param_and_grads):`
`152`	`152`
`153`	`153`	`def apply_optimize(self, loss, startup_program, params_grads):`
`154`	`154`	`new_params_grads = self.fp16_compression(params_grads)`
`155`		`- return self.inner_opt.apply_optimize(`
	`155`	`+ return self.inner_opt._apply_optimize(`
`156`	`156`	`loss, startup_program=startup_program, params_grads=new_params_grads`
`157`	`157`	`)`
Original file line number	Diff line number	Diff line change
`@@ -49,9 +49,7 @@ def _can_apply(self):`
`49`	`49`	`self.inner_opt,`
`50`	`50`	`(`
`51`	`51`	`paddle.optimizer.momentum.Momentum,`
`52`		`- paddle.fluid.optimizer.Momentum,`
`53`	`52`	`paddle.optimizer.sgd.SGD,`
`54`		`- paddle.fluid.optimizer.SGD,`
`55`	`53`	`),`
`56`	`54`	`)`
`57`	`55`
`@@ -235,9 +233,7 @@ def _can_apply(self):`
`235`	`233`	`self.inner_opt,`
`236`	`234`	`(`
`237`	`235`	`paddle.optimizer.Momentum,`
`238`		`- paddle.fluid.optimizer.Momentum,`
`239`	`236`	`paddle.optimizer.sgd.SGD,`
`240`		`- paddle.fluid.optimizer.SGD,`
`241`	`237`	`),`
`242`	`238`	`)`
`243`	`239`
Original file line number	Diff line number	Diff line change
`@@ -12,7 +12,7 @@`
`12`	`12`	`# See the License for the specific language governing permissions and`
`13`	`13`	`# limitations under the License.`
`14`	`14`
`15`		`-from paddle.fluid.optimizer import Optimizer`
	`15`	`+from paddle.optimizer import Optimizer`
`16`	`16`
`17`	`17`	`__all__ = []`
`18`	`18`
`@@ -81,7 +81,7 @@ def backward(`
`81`	`81`	`)`
`82`	`82`
`83`	`83`	`def apply_optimize(self, loss, startup_program, params_grads):`
`84`		`- return self.inner_opt.apply_optimize(`
	`84`	`+ return self.inner_opt._apply_optimize(`
`85`	`85`	`loss, startup_program=startup_program, params_grads=params_grads`
`86`	`86`	`)`
`87`	`87`