fix bug in saving model (invoke the wrong function)

01.fit_a_line/train.py

@@ -41,7 +41,7 @@ def main():
         if isinstance(event, paddle.event.EndPass):
             if event.pass_id % 10 == 0:
                 with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
-                    parameters.to_tar(f)
+                    trainer.save_parameter_to_tar(f)
             result = trainer.test(
                 reader=paddle.batch(uci_housing.test(), batch_size=2),
                 feeding=feeding)

02.recognize_digits/train.py

@@ -87,7 +87,7 @@ def main():
         if isinstance(event, paddle.event.EndPass):
             # save parameters
             with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
-                parameters.to_tar(f)
+                trainer.save_parameter_to_tar(f)
 
             result = trainer.test(reader=paddle.batch(
                 paddle.dataset.mnist.test(), batch_size=128))

03.image_classification/train.py

@@ -57,6 +57,10 @@ def main():
         learning_rate_decay_b=50000 * 100,
         learning_rate_schedule='discexp')
 
+    # Create trainer
+    trainer = paddle.trainer.SGD(
+        cost=cost, parameters=parameters, update_equation=momentum_optimizer)
+
     # End batch and end pass event handler
     def event_handler(event):
         if isinstance(event, paddle.event.EndIteration):
@@ -69,7 +73,7 @@ def main():
         if isinstance(event, paddle.event.EndPass):
             # save parameters
             with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
-                parameters.to_tar(f)
+                trainer.save_parameter_to_tar(f)
 
             result = trainer.test(
                 reader=paddle.batch(
@@ -78,10 +82,6 @@ def main():
                          'label': 1})
             print "\nTest with Pass %d, %s" % (event.pass_id, result.metrics)
 
-    # Create trainer
-    trainer = paddle.trainer.SGD(
-        cost=cost, parameters=parameters, update_equation=momentum_optimizer)
-
     # Save the inference topology to protobuf.
     inference_topology = paddle.topology.Topology(layers=out)
     with open("inference_topology.pkl", 'wb') as f:

04.word2vec/train.py

@@ -76,6 +76,14 @@ def main():
         bias_attr=paddle.attr.Param(learning_rate=2),
         act=paddle.activation.Softmax())
 
+
+    cost = paddle.layer.classification_cost(input=predictword, label=nextword)
+    parameters = paddle.parameters.create(cost)
+    adagrad = paddle.optimizer.AdaGrad(
+        learning_rate=3e-3,
+        regularization=paddle.optimizer.L2Regularization(8e-4))
+    trainer = paddle.trainer.SGD(cost, parameters, adagrad)
+
     def event_handler(event):
         if isinstance(event, paddle.event.EndIteration):
             if event.batch_id % 100 == 0:
@@ -88,14 +96,7 @@ def main():
             print "Pass %d, Testing metrics %s" % (event.pass_id,
                                                    result.metrics)
             with open("model_%d.tar" % event.pass_id, 'w') as f:
-                parameters.to_tar(f)
-
-    cost = paddle.layer.classification_cost(input=predictword, label=nextword)
-    parameters = paddle.parameters.create(cost)
-    adagrad = paddle.optimizer.AdaGrad(
-        learning_rate=3e-3,
-        regularization=paddle.optimizer.L2Regularization(8e-4))
-    trainer = paddle.trainer.SGD(cost, parameters, adagrad)
+                trainer.save_parameter_to_tar(f)
     trainer.train(
         paddle.batch(paddle.dataset.imikolov.train(word_dict, N), 32),
         num_passes=100,

06.understand_sentiment/train.py

@@ -129,6 +129,10 @@ if __name__ == '__main__':
         regularization=paddle.optimizer.L2Regularization(rate=8e-4),
         model_average=paddle.optimizer.ModelAverage(average_window=0.5))
 
+    # create trainer
+    trainer = paddle.trainer.SGD(
+        cost=cost, parameters=parameters, update_equation=adam_optimizer)
+
     # End batch and end pass event handler
     def event_handler(event):
         if isinstance(event, paddle.event.EndIteration):
@@ -140,14 +144,11 @@ if __name__ == '__main__':
                 sys.stdout.flush()
         if isinstance(event, paddle.event.EndPass):
             with open('./params_pass_%d.tar' % event.pass_id, 'w') as f:
-                parameters.to_tar(f)
+                trainer.save_parameter_to_tar(f)
 
             result = trainer.test(reader=test_reader, feeding=feeding)
             print "\nTest with Pass %d, %s" % (event.pass_id, result.metrics)
 
-    # create trainer
-    trainer = paddle.trainer.SGD(
-        cost=cost, parameters=parameters, update_equation=adam_optimizer)
     # Save the inference topology to protobuf.
     inference_topology = paddle.topology.Topology(layers=output)
     with open("./inference_topology.pkl", 'wb') as f:

07.label_semantic_roles/train.py

@@ -188,7 +188,7 @@ def main():
         if isinstance(event, paddle.event.EndPass):
             # save parameters
             with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
-                parameters.to_tar(f)
+                trainer.save_parameter_to_tar(f)
 
             result = trainer.test(reader=test_reader, feeding=feeding)
             print "\nTest with Pass %d, %s" % (event.pass_id, result.metrics)

08.machine_translation/train.py

@@ -5,9 +5,9 @@ import paddle.v2 as paddle
 with_gpu = os.getenv('WITH_GPU', '0') != '0'
 
 
-def save_model(parameters, save_path):
+def save_model(trainer, parameters, save_path):
     with open(save_path, 'w') as f:
-        parameters.to_tar(f)
+        trainer.save_parameter_to_tar(f)
 
 
 def seq_to_seq_net(source_dict_dim,
@@ -175,12 +175,12 @@ def main():
                 if not event.batch_id % 10:
                     save_path = 'params_pass_%05d_batch_%05d.tar' % (
                         event.pass_id, event.batch_id)
-                    save_model(parameters, save_path)
+                    save_model(trainer, parameters, save_path)
 
             if isinstance(event, paddle.event.EndPass):
                 # save parameters
                 save_path = 'params_pass_%05d.tar' % (event.pass_id)
-                save_model(parameters, save_path)
+                save_model(trainer, parameters, save_path)
 
         # start to train
         trainer.train(