[remove fluid.layers.cross_entropy] remove unit tests (part 2)

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

@@ -806,7 +806,12 @@ def _test(
                     input=fc_1, size=2, param_attr=weight_attr2, 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,
+                )
                 loss = paddle.mean(cost)
                 beta1_init = 0.9
                 beta2_init = 0.999
@@ -966,7 +971,9 @@ def test_adam_exception(self):
         fc_1 = fluid.layers.fc(input=z, size=128)
         prediction = fluid.layers.fc(input=fc_1, size=2, 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
+        )
         loss = paddle.mean(cost)
         adam = fluid.optimizer.Adam(use_global_beta_pow=True)
         adam.minimize(loss)

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

@@ -57,7 +57,9 @@ def convolutional_neural_network(use_py_reader):
         )
 
         prediction = fluid.layers.fc(input=conv_pool_2, size=10, 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
+        )
         avg_loss = paddle.mean(loss)
         acc = paddle.static.accuracy(input=prediction, label=label)
         i = fluid.layers.zeros(shape=[1], dtype='int64')

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

@@ -107,7 +107,9 @@ def build_simple_model(self):
         )
         label = fluid.layers.data(name='label', shape=[1], dtype='int64')
         prediction = fluid.layers.fc(input=img, size=10, 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
+        )
         avg_loss = paddle.mean(loss)
 
     def compile_program_not_compiled(self):

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

@@ -17,6 +17,7 @@
 import numpy as np
 from op_test import OpTest, randomize_probability
 
+import paddle
 import paddle.fluid as fluid
 import paddle.fluid.core as core
 from paddle.fluid import Program, program_guard
@@ -419,7 +420,9 @@ def test_Variable():
                 lab1 = fluid.create_lod_tensor(
                     np.array([-1, 3, 5, 5]), [[1, 1, 1, 1]], fluid.CPUPlace()
                 )
-                fluid.layers.cross_entropy(x1, lab1)
+                paddle.nn.functional.cross_entropy(
+                    x1, lab1, reduction='none', use_softmax=False
+                )
 
             self.assertRaises(TypeError, test_Variable)
 
@@ -432,7 +435,9 @@ def test_dtype():
                 lab2 = fluid.layers.data(
                     name='lab2', shape=[3, 4, 5, 6], dtype="int32"
                 )
-                fluid.layers.cross_entropy(x2, lab2)
+                paddle.nn.functional.cross_entropy(
+                    x2, lab2, reduction='none', use_softmax=False
+                )
 
             self.assertRaises(TypeError, test_dtype)
 

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

@@ -67,7 +67,12 @@ def simple_fc_net(places, use_legacy_py_reader, use_double_buffer):
                 hidden, size=CLASS_NUM, act='softmax'
             )
             loss = paddle.mean(
-                fluid.layers.cross_entropy(input=predict_label, label=label)
+                paddle.nn.functional.cross_entropy(
+                    input=predict_label,
+                    label=label,
+                    reduction='none',
+                    use_softmax=False,
+                )
             )
 
             optimizer = fluid.optimizer.Adam()

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

@@ -73,7 +73,9 @@ def get_model(batch_size):
 
     # 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
@@ -188,9 +190,11 @@ def test_clone_with_stop_gradient(self):
             hidden1 = fluid.layers.fc(input=img, size=200, act='relu')
             hidden1.stop_gradient = True
             hidden2 = fluid.layers.dropout(hidden1, dropout_prob=0.5)
-            loss = fluid.layers.cross_entropy(
+            loss = paddle.nn.functional.cross_entropy(
                 input=fluid.layers.fc(hidden2, size=10, act='softmax'),
                 label=fluid.layers.data(name='label', shape=[1], dtype='int64'),
+                reduction='none',
+                use_softmax=False,
             )
             avg_loss = paddle.mean(loss)
             test_program = train_program.clone(for_test=False)
@@ -226,9 +230,11 @@ def false_fn():
 
             hidden2 = fluid.layers.cond(cond, true_fn, false_fn)
 
-            loss = fluid.layers.cross_entropy(
+            loss = paddle.nn.functional.cross_entropy(
                 input=fluid.layers.fc(hidden2, size=10, act='softmax'),
                 label=fluid.layers.data(name='label', shape=[1], dtype='int64'),
+                reduction='none',
+                use_softmax=False,
             )
             avg_loss = paddle.mean(loss)
             test_program = train_program.clone(for_test=False)
@@ -266,9 +272,11 @@ def false_fn():
                 return hidden2
 
             hidden2 = fluid.layers.cond(cond, true_fn, false_fn)
-            loss = fluid.layers.cross_entropy(
+            loss = paddle.nn.functional.cross_entropy(
                 input=fluid.layers.fc(hidden2, size=10, act='softmax'),
                 label=fluid.layers.data(name='label', shape=[1], dtype='int64'),
+                reduction='none',
+                use_softmax=False,
             )
             avg_loss = paddle.mean(loss)
             test_program = train_program.clone(for_test=False)

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

@@ -52,8 +52,8 @@ def test_a_sync_optimizer1(self):
         fc_1 = paddle.fluid.layers.fc(input=input_x, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
 

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

@@ -68,8 +68,8 @@ def test_a_sync_optimizer3(self):
         fc_1 = paddle.fluid.layers.fc(input=x_embedding, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
 

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

@@ -56,8 +56,8 @@ def test_a_sync_optimizer2(self):
         fc_1 = paddle.fluid.layers.fc(input=emb, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
         os.environ["FLAGS_LAUNCH_BARRIER"] = "0"

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

@@ -51,8 +51,8 @@ def test_a_sync_optimizer_trainer(self):
         fc_1 = paddle.fluid.layers.fc(input=input_x, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
 
@@ -84,8 +84,8 @@ def test_a_sync_optimizer_pserver(self):
         fc_1 = paddle.fluid.layers.fc(input=input_x, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
 

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

@@ -150,7 +150,9 @@ def embedding_layer(input):
 
         with fluid.device_guard("gpu"):
             labels = fluid.layers.cast(inputs[-1], dtype="int64")
-            cost = fluid.layers.cross_entropy(input=predict, label=labels)
+            cost = paddle.nn.functional.cross_entropy(
+                input=predict, label=labels, reduction='none', use_softmax=False
+            )
             avg_cost = paddle.sum(cost)
 
         return avg_cost

python/paddle/fluid/tests/unittests/test_dygraph_mnist_fp16.py

@@ -112,7 +112,9 @@ def forward(self, inputs, label):
         x = paddle.reshape(x, shape=[-1, self.pool_2_shape])
         cost = self._linear(x)
         cost = paddle.nn.functional.softmax(cost)
-        loss = fluid.layers.cross_entropy(cost, label)
+        loss = paddle.nn.functional.cross_entropy(
+            cost, label, reduction='none', use_softmax=False
+        )
         avg_loss = paddle.mean(loss)
         return avg_loss
 

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

@@ -138,7 +138,9 @@ def test_mnist_forward_float32(self):
                     label.stop_gradient = True
 
                     cost = mnist(img)
-                    loss = fluid.layers.cross_entropy(cost, label)
+                    loss = paddle.nn.functional.cross_entropy(
+                        cost, label, reduction='none', use_softmax=False
+                    )
                     avg_loss = paddle.mean(loss)
 
                     dy_out = avg_loss.numpy()
@@ -167,7 +169,9 @@ def test_mnist_forward_float32(self):
             )
             label = fluid.layers.data(name='label', shape=[1], dtype='int64')
             cost = mnist(img)
-            loss = fluid.layers.cross_entropy(cost, label)
+            loss = paddle.nn.functional.cross_entropy(
+                cost, label, reduction='none', use_softmax=False
+            )
             avg_loss = paddle.mean(loss)
 
             # initialize params and fetch them

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

@@ -45,7 +45,9 @@ def simple_fc_net():
             ),
         )
     prediction = fluid.layers.fc(hidden, size=10, 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)
     optimizer = fluid.optimizer.Adam(learning_rate=1e-3)
     optimizer.minimize(loss)

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

@@ -43,7 +43,9 @@ def gru_net(
     gru_max_tanh = paddle.tanh(gru_max)
     fc1 = fluid.layers.fc(input=gru_max_tanh, size=hid_dim2, act='tanh')
     prediction = fluid.layers.fc(input=fc1, 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/test_eager_deletion_lstm_net.py

@@ -45,7 +45,9 @@ def lstm_net(
     lstm_max_tanh = paddle.tanh(lstm_max)
     fc1 = fluid.layers.fc(input=lstm_max_tanh, size=hid_dim2, act='tanh')
     prediction = fluid.layers.fc(input=fc1, 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/test_feed_data_check_shape_type.py

@@ -67,7 +67,12 @@ def _simple_fc_net(self, in_size, label_size, class_num, hidden_sizes):
 
         predict_label = fluid.layers.fc(hidden, size=class_num, act='softmax')
         loss = paddle.mean(
-            fluid.layers.cross_entropy(input=predict_label, label=label)
+            paddle.nn.functional.cross_entropy(
+                input=predict_label,
+                label=label,
+                reduction='none',
+                use_softmax=False,
+            )
         )
 
         optimizer = fluid.optimizer.Adam()

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

@@ -46,7 +46,9 @@ def conv_net(self, img, label):
         )
         hidden = fluid.layers.fc(input=conv_pool_2, size=32, act='relu')
         prediction = fluid.layers.fc(input=hidden, size=10, 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
+        )
         avg_loss = paddle.mean(loss)
         return avg_loss, prediction
 

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

@@ -40,8 +40,8 @@ def test_distributed_strategy_auto(self):
         fc_1 = paddle.fluid.layers.fc(input=input_x, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
 

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

@@ -203,7 +203,12 @@ def test_single_error():
 
             fc_1 = fluid.layers.fc(input=input_x, size=64, act='tanh')
             prediction = fluid.layers.fc(input=fc_1, size=2, act='softmax')
-            cost = fluid.layers.cross_entropy(input=prediction, label=input_y)
+            cost = paddle.nn.functional.cross_entropy(
+                input=prediction,
+                label=input_y,
+                reduction='none',
+                use_softmax=False,
+            )
             avg_cost = paddle.mean(x=cost)
             fleet.init(is_collective=True)
 

python/paddle/fluid/tests/unittests/test_fleet_base_2.py

@@ -54,8 +54,8 @@ def test_ps_minimize(self):
         fc_1 = paddle.fluid.layers.fc(input=input_x, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
 

python/paddle/fluid/tests/unittests/test_fleet_base_3.py

@@ -40,8 +40,8 @@ def test_collective_minimize(self):
         fc_1 = paddle.fluid.layers.fc(input=input_x, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
 
@@ -71,8 +71,8 @@ def test_fleet_get_applied_optimizer(self):
         fc_1 = paddle.fluid.layers.fc(input=input_x, size=64, act='tanh')
         fc_2 = paddle.fluid.layers.fc(input=fc_1, size=64, act='tanh')
         prediction = paddle.fluid.layers.fc(input=[fc_2], size=2, act='softmax')
-        cost = paddle.fluid.layers.cross_entropy(
-            input=prediction, label=input_y
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
         )
         avg_cost = paddle.mean(x=cost)
 

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

@@ -85,7 +85,9 @@ def test_single_run_collective_minimize(self):
 
         fc_1 = fluid.layers.fc(input=input_x, size=64, act='tanh')
         prediction = fluid.layers.fc(input=fc_1, size=2, act='softmax')
-        cost = fluid.layers.cross_entropy(input=prediction, label=input_y)
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         fleet.init(is_collective=True)
@@ -124,7 +126,9 @@ def test_single_run_ps_minimize(self):
 
         fc_1 = fluid.layers.fc(input=input_x, size=64, act='tanh')
         prediction = fluid.layers.fc(input=fc_1, size=2, act='softmax')
-        cost = fluid.layers.cross_entropy(input=prediction, label=input_y)
+        cost = paddle.nn.functional.cross_entropy(
+            input=prediction, label=input_y, reduction='none', use_softmax=False
+        )
         avg_cost = paddle.mean(x=cost)
 
         fleet.init()

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

@@ -53,7 +53,9 @@ def build_program(self, main_program, startup_program, use_cuda, seed=1):
                 input=hidden3, act='relu', data_layout='NHWC'
             )
             prediction = fluid.layers.fc(input=hidden4, size=10, act='softmax')
-            loss = fluid.layers.cross_entropy(input=prediction, label=y)
+            loss = paddle.nn.functional.cross_entropy(
+                input=prediction, label=y, reduction='none', use_softmax=False
+            )
             loss = paddle.mean(loss)
             sgd = fluid.optimizer.SGD(learning_rate=0.001)
             if use_cuda: