use new save load in all dygraph modell (PaddlePaddle#3495)

* user new save load in all dygraph modell; test=develop * change load_dict to set_dict; test=develop
nuaaceieyty · Oct 10, 2019 · 7a3e0c7 · 7a3e0c7
1 parent 7e59194
commit 7a3e0c7
Show file tree

Hide file tree

Showing 10 changed files with 18 additions and 18 deletions.
diff --git a/dygraph/cycle_gan/infer.py b/dygraph/cycle_gan/infer.py
@@ -52,8 +52,8 @@ def infer():
             os.makedirs(out_path)
         cycle_gan = Cycle_Gan("cycle_gan")
         save_dir = args.init_model 
-        restore, _ = fluid.dygraph.load_persistables(save_dir)
-        cycle_gan.load_dict(restore)
+        restore, _ = fluid.load_dygraph(save_dir)
+        cycle_gan.set_dict(restore)
         cycle_gan.eval()
         for file in glob.glob(args.input):
             print ("read %s" % file)

diff --git a/dygraph/cycle_gan/test.py b/dygraph/cycle_gan/test.py
@@ -52,8 +52,8 @@ def test():
             os.makedirs(out_path)
         cycle_gan = Cycle_Gan("cycle_gan")
         save_dir = args.init_model + str(epoch)
-        restore, _ = fluid.dygraph.load_persistables(save_dir)
-        cycle_gan.load_dict(restore)
+        restore, _ = fluid.load_dygraph(save_dir)
+        cycle_gan.set_dict(restore)
         cycle_gan.eval()
         for data_A , data_B in zip(A_test_reader(), B_test_reader()): 
             A_name = data_A[1]

diff --git a/dygraph/cycle_gan/train.py b/dygraph/cycle_gan/train.py
@@ -204,7 +204,7 @@ def train(args):
                     break
 
             if args.save_checkpoints:
-                fluid.dygraph.save_persistables(
+                fluid.save_dygraph(
                     cycle_gan.state_dict(),
                     args.output + "/checkpoints/{}".format(epoch))
 

diff --git a/dygraph/mnist/train.py b/dygraph/mnist/train.py
@@ -150,8 +150,8 @@ def inference_mnist():
     with fluid.dygraph.guard(place):
         mnist_infer = MNIST("mnist")
         # load checkpoint
-        model_dict, _ = fluid.dygraph.load_persistables("save_dir")
-        mnist_infer.load_dict(model_dict)
+        model_dict, _ = fluid.load_dygraph("save_temp")
+        mnist_infer.set_dict(model_dict)
         print("checkpoint loaded")
 
         # start evaluate mode
@@ -245,7 +245,7 @@ def train_mnist(args):
             args.use_data_parallel and
             fluid.dygraph.parallel.Env().local_rank == 0)
         if save_parameters:
-            fluid.dygraph.save_persistables(mnist.state_dict(), "save_dir")
+            fluid.save_dygraph(mnist.state_dict(), "save_temp")
             print("checkpoint saved")
 
         inference_mnist()

diff --git a/dygraph/reinforcement_learning/actor_critic.py b/dygraph/reinforcement_learning/actor_critic.py
@@ -200,5 +200,5 @@ def finish_episode():
             print("Solved! Running reward is now {} and "
                   "the last episode runs to {} time steps!".format(
                       running_reward, t))
-            fluid.dygraph.save_persistables(policy.state_dict(), args.save_dir)
+            fluid.save_dygraph(policy.state_dict(), args.save_dir)
             break
diff --git a/dygraph/reinforcement_learning/reinforce.py b/dygraph/reinforcement_learning/reinforce.py
@@ -186,5 +186,5 @@ def finish_episode():
             print("Solved! Running reward is now {} and "
                   "the last episode runs to {} time steps!".format(
                       running_reward, t))
-            fluid.dygraph.save_persistables(policy.state_dict(), args.save_dir)
+            fluid.save_dygraph(policy.state_dict(), args.save_dir)
             break
diff --git a/dygraph/reinforcement_learning/test_actor_critic_load.py b/dygraph/reinforcement_learning/test_actor_critic_load.py
@@ -174,8 +174,8 @@ def finish_episode():
         return returns
 
     running_reward = 10
-    model_dict, _ = fluid.dygraph.load_persistables(args.save_dir)
-    policy.load_dict(model_dict)
+    model_dict, _ = fluid.load_dygraph(args.save_dir)
+    policy.set_dict(model_dict)
 
     state, ep_reward = env.reset(), 0
     for t in range(1, 10000):  # Don't infinite loop while learning

diff --git a/dygraph/reinforcement_learning/test_reinforce_load.py b/dygraph/reinforcement_learning/test_reinforce_load.py
@@ -161,8 +161,8 @@ def finish_episode():
 
     running_reward = 10
     state, ep_reward = env.reset(), 0
-    model_dict, _ = fluid.dygraph.load_persistables(args.save_dir)
-    policy.load_dict(model_dict)
+    model_dict, _ = fluid.load_dygraph(args.save_dir)
+    policy.set_dict(model_dict)
 
     for t in range(1, 10000):  # Don't infinite loop while learning
         state = np.array(state).astype("float32")

diff --git a/dygraph/resnet/train.py b/dygraph/resnet/train.py
@@ -380,7 +380,7 @@ def train_resnet():
                 args.use_data_parallel and
                 fluid.dygraph.parallel.Env().local_rank == 0)
             if save_parameters:
-                fluid.dygraph.save_persistables(resnet.state_dict(),
+                fluid.save_dygraph(resnet.state_dict(),
                                                 'resnet_params')
 
 

diff --git a/dygraph/sentiment/main.py b/dygraph/sentiment/main.py
@@ -232,7 +232,7 @@ def train():
                     if steps % args.save_steps == 0:
                         save_path = "save_dir_" + str(steps)
                         print('save model to: ' + save_path)
-                        fluid.dygraph.save_persistables(cnn_net.state_dict(),
+                        fluid.save_dygraph(cnn_net.state_dict(),
                                                         save_path)
                 if enable_profile:
                         print('save profile result into /tmp/profile_file')
@@ -258,8 +258,8 @@ def infer():
         print('Do inferring ...... ')
         total_acc, total_num_seqs = [], []
 
-        restore, _ = fluid.dygraph.load_persistables(args.checkpoints)
-        cnn_net_infer.load_dict(restore)
+        restore, _ = fluid.load_dygraph(args.checkpoints)
+        cnn_net_infer.set_dict(restore)
         cnn_net_infer.eval()
 
         steps = 0