[remove fluid.layers.cross_entropy] remove unit tests (part 4)

python/paddle/fluid/tests/unittests/dist_allreduce_op.py

@@ -77,7 +77,9 @@ def get_model(self, batch_size=2, single_device=False):
 
         # Train program
         predict = cnn_model(images)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         # Evaluator

python/paddle/fluid/tests/unittests/dist_ctr.py

@@ -103,7 +103,9 @@ def get_model(self, batch_size=2):
         auc_var, batch_auc_var, auc_states = paddle.static.auc(
             input=predict, label=label
         )
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         inference_program = paddle.fluid.default_main_program().clone()

python/paddle/fluid/tests/unittests/dist_fleet_ctr.py

@@ -154,7 +154,9 @@ def net(self, args, is_train=True, batch_size=4, lr=0.01):
             input=predict, label=label
         )
 
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         self.feeds = datas

python/paddle/fluid/tests/unittests/dist_fleet_heter_pipeline_ctr.py

@@ -123,7 +123,9 @@ def net(self, args, batch_size=4, lr=0.01):
             label = fluid.layers.cast(label, dtype="int64")
             predict = fluid.layers.fc(input=merge_layer, size=2, act='softmax')
 
-            cost = fluid.layers.cross_entropy(input=predict, label=label)
+            cost = paddle.nn.functional.cross_entropy(
+                input=predict, label=label, reduction='none', use_softmax=False
+            )
             avg_cost = paddle.mean(x=cost)
             fluid.layers.Print(avg_cost, message="avg_cost")
 

python/paddle/fluid/tests/unittests/dist_fleet_raw_program_optimizer.py

@@ -79,7 +79,9 @@ def get_model(self, batch_size=2, single_device=False):
 
         # Train program
         predict = cnn_model(images)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         # Evaluator

python/paddle/fluid/tests/unittests/dist_fleet_raw_program_optimizer_fuse_allreduce.py

@@ -79,7 +79,9 @@ def get_model(self, batch_size=2, single_device=False):
 
         # Train program
         predict = cnn_model(images)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         # Evaluator

python/paddle/fluid/tests/unittests/dist_fleet_sparse_embedding_ctr.py

@@ -140,7 +140,9 @@ def net(self, args, batch_size=4, lr=0.01):
 
         acc = paddle.static.accuracy(input=predict, label=label)
         auc_var, _, _ = paddle.static.auc(input=predict, label=label)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         self.feeds = datas

python/paddle/fluid/tests/unittests/dist_mnist.py

@@ -78,7 +78,9 @@ def get_model(self, batch_size=2, use_dgc=False, dist_strategy=None):
 
         # Train program
         predict = cnn_model(images)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         # Evaluator

python/paddle/fluid/tests/unittests/dist_mnist_batch_merge.py

@@ -43,7 +43,9 @@ def get_model(self, batch_size=2):
 
         # Train program
         predict = cnn_model(images)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         # Evaluator

python/paddle/fluid/tests/unittests/dist_mnist_fp16_allreduce.py

@@ -37,7 +37,9 @@ def get_model(self, batch_size=2):
 
         # Train program
         predict = cnn_model(images)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         # Evaluator

python/paddle/fluid/tests/unittests/dist_mnist_lars.py

@@ -34,7 +34,9 @@ def get_model(self, batch_size=2):
 
         # Train program
         predict = cnn_model(images)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         # Evaluator

python/paddle/fluid/tests/unittests/dist_se_resnext.py

@@ -221,7 +221,9 @@ def get_model(self, batch_size=2, use_dgc=False):
         # Train program
         model = SE_ResNeXt(layers=50)
         out = model.net(input=image, class_dim=102)
-        cost = fluid.layers.cross_entropy(input=out, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=out, label=label, reduction='none', use_softmax=False
+        )
 
         avg_cost = paddle.mean(x=cost)
         acc_top1 = paddle.static.accuracy(input=out, label=label, k=1)

python/paddle/fluid/tests/unittests/dist_text_classification.py

@@ -132,7 +132,9 @@ def get_model(self, batch_size=2):
 
         # Train program
         predict = conv_net(data, dict_dim)
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
         acc = paddle.static.accuracy(input=predict, label=label)
         inference_program = fluid.default_main_program().clone()

python/paddle/fluid/tests/unittests/fleet_heter_ps_training.py

@@ -112,7 +112,9 @@ def net(batch_size=4, lr=0.01):
         label = fluid.layers.cast(label, dtype="int64")
         predict = fluid.layers.fc(input=merge_layer, size=2, act='softmax')
 
-        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        cost = paddle.nn.functional.cross_entropy(
+            input=predict, label=label, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
     return datas, avg_cost
 

python/paddle/fluid/tests/unittests/parallel_dygraph_mnist.py

@@ -95,7 +95,9 @@ def forward(self, inputs, label):
         x = self._simple_img_conv_pool_2(x)
         x = paddle.reshape(x, shape=[-1, self.pool_2_shape])
         cost = self._fc(x)
-        loss = fluid.layers.cross_entropy(self.act(cost), label)
+        loss = paddle.nn.functional.cross_entropy(
+            self.act(cost), label, reduction='none', use_softmax=False
+        )
         avg_loss = paddle.mean(loss)
         return avg_loss
 

python/paddle/fluid/tests/unittests/seresnext_net.py

@@ -170,7 +170,9 @@ def SE_ResNeXt50Small(use_feed):
     )
     # Classifier layer:
     prediction = fluid.layers.fc(input=dropout, size=1000, act='softmax')
-    loss = fluid.layers.cross_entropy(input=prediction, label=label)
+    loss = paddle.nn.functional.cross_entropy(
+        input=prediction, label=label, reduction='none', use_softmax=False
+    )
     loss = paddle.mean(loss)
     return loss
 

python/paddle/fluid/tests/unittests/simple_nets.py

@@ -30,7 +30,9 @@ def simple_fc_net_with_inputs(img, label, class_num=10):
             ),
         )
     prediction = fluid.layers.fc(hidden, size=class_num, act='softmax')
-    loss = fluid.layers.cross_entropy(input=prediction, label=label)
+    loss = paddle.nn.functional.cross_entropy(
+        input=prediction, label=label, reduction='none', use_softmax=False
+    )
     loss = paddle.mean(loss)
     return loss
 
@@ -56,7 +58,9 @@ def batchnorm_fc_with_inputs(img, label, class_num=10):
         hidden = paddle.static.nn.batch_norm(input=hidden)
 
     prediction = fluid.layers.fc(hidden, size=class_num, act='softmax')
-    loss = fluid.layers.cross_entropy(input=prediction, label=label)
+    loss = paddle.nn.functional.cross_entropy(
+        input=prediction, label=label, reduction='none', use_softmax=False
+    )
     loss = paddle.mean(loss)
     return loss
 
@@ -93,7 +97,9 @@ def bow_net(
     fc_1 = fluid.layers.fc(input=bow_tanh, size=hid_dim, act="tanh")
     fc_2 = fluid.layers.fc(input=fc_1, size=hid_dim2, act="tanh")
     prediction = fluid.layers.fc(input=[fc_2], size=class_dim, act="softmax")
-    cost = fluid.layers.cross_entropy(input=prediction, label=label)
+    cost = paddle.nn.functional.cross_entropy(
+        input=prediction, label=label, reduction='none', use_softmax=False
+    )
     avg_cost = paddle.mean(x=cost)
 
     return avg_cost

python/paddle/fluid/tests/unittests/transformer_model.py

@@ -594,6 +594,8 @@ def transformer(
     )
     predict = paddle.nn.functional.softmax(predict)
 
-    cost = layers.cross_entropy(input=predict, label=gold)
+    cost = paddle.nn.functional.cross_entropy(
+        input=predict, label=gold, reduction='none', use_softmax=False
+    )
     weighted_cost = cost * weights
     return paddle.sum(weighted_cost)