clean elem_arithmetic part5 unittest (#48466)

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

@@ -39,7 +39,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -49,13 +49,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

python/paddle/fluid/tests/unittests/test_dist_fleet_ps11.py

@@ -39,7 +39,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -49,13 +49,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

python/paddle/fluid/tests/unittests/test_dist_fleet_ps12.py

@@ -42,7 +42,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -52,13 +52,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

python/paddle/fluid/tests/unittests/test_dist_fleet_ps13.py

@@ -43,7 +43,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -53,13 +53,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

python/paddle/fluid/tests/unittests/test_dist_fleet_ps2.py

@@ -42,7 +42,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -52,13 +52,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

python/paddle/fluid/tests/unittests/test_dist_fleet_ps3.py

@@ -39,7 +39,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -49,13 +49,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

python/paddle/fluid/tests/unittests/test_dist_fleet_ps4.py

@@ -39,7 +39,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -49,13 +49,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

python/paddle/fluid/tests/unittests/test_dist_fleet_ps5.py

@@ -39,7 +39,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -49,13 +49,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

python/paddle/fluid/tests/unittests/test_dist_fleet_ps6.py

@@ -39,7 +39,7 @@ class TestPSPassWithBow(unittest.TestCase):
             cond = fluid.layers.less_than(cos_q_nt, cos_q_pt)
             cond = fluid.layers.cast(cond, dtype='float64')
             cond_3 = paddle.sum(cond)
-            acc = fluid.layers.elementwise_div(
+            acc = paddle.divide(
                 cond_3,
                 fluid.layers.fill_constant(
                     shape=[1], value=batch_size * 1.0, dtype='float64'
@@ -49,13 +49,13 @@ class TestPSPassWithBow(unittest.TestCase):
             return acc
 
         def get_loss(cos_q_pt, cos_q_nt):
-            loss_op1 = fluid.layers.elementwise_sub(
+            loss_op1 = paddle.subtract(
                 fluid.layers.fill_constant_batch_size_like(
                     input=cos_q_pt, shape=[-1, 1], value=margin, dtype='float32'
                 ),
                 cos_q_pt,
             )
-            loss_op2 = fluid.layers.elementwise_add(loss_op1, cos_q_nt)
+            loss_op2 = paddle.add(loss_op1, cos_q_nt)
             loss_op3 = paddle.maximum(
                 fluid.layers.fill_constant_batch_size_like(
                     input=loss_op2, shape=[-1, 1], value=0.0, dtype='float32'

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

@@ -403,9 +403,9 @@ class TestFakeInit(TranspilerTest):
 
         neg_emb_b_vec = paddle.reshape(neg_emb_b, shape=[-1, neg_num])
 
-        true_logits = fluid.layers.elementwise_add(
+        true_logits = paddle.add(
             paddle.sum(
-                fluid.layers.elementwise_mul(input_emb, true_emb_w),
+                paddle.multiply(input_emb, true_emb_w),
                 dim=1,
                 keep_dim=True,
             ),
@@ -418,7 +418,7 @@ class TestFakeInit(TranspilerTest):
             input_emb_re, neg_emb_w_re, transpose_y=True
         )
         neg_matmul_re = paddle.reshape(neg_matmul, shape=[-1, neg_num])
-        neg_logits = fluid.layers.elementwise_add(neg_matmul_re, neg_emb_b_vec)
+        neg_logits = paddle.add(neg_matmul_re, neg_emb_b_vec)
         # nce loss
         label_ones = fluid.layers.fill_constant_batch_size_like(
             true_logits, shape=[-1, 1], value=1.0, dtype='float32'
@@ -433,7 +433,7 @@ class TestFakeInit(TranspilerTest):
         neg_xent = paddle.nn.functional.binary_cross_entropy_with_logits(
             neg_logits, label_zeros
         )
-        cost = fluid.layers.elementwise_add(
+        cost = paddle.add(
             paddle.sum(true_xent, axis=1),
             paddle.sum(neg_xent, axis=1),
         )

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

@@ -169,7 +169,7 @@ def lm_model(
                 nn = layers.concat([input, pre_hidden], 1)
                 gate_input = layers.matmul(x=nn, y=weight_1)
 
-                gate_input = layers.elementwise_add(gate_input, bias)
+                gate_input = paddle.add(gate_input, bias)
                 i = paddle.slice(
                     gate_input, axes=[1], starts=[0], ends=[hidden_size]
                 )
@@ -293,7 +293,7 @@ def lm_model(
                 nn = layers.concat([input, pre_hidden], 1)
                 gate_input = layers.matmul(x=nn, y=weight_1)
 
-                gate_input = layers.elementwise_add(gate_input, bias)
+                gate_input = paddle.add(gate_input, bias)
                 i, j, f, o = layers.split(gate_input, num_or_sections=4, dim=-1)
 
                 c = pre_cell * paddle.nn.functional.sigmoid(
@@ -460,7 +460,7 @@ def lm_model(
     )
 
     projection = layers.matmul(rnn_out, softmax_weight)
-    projection = layers.elementwise_add(projection, softmax_bias)
+    projection = paddle.add(projection, softmax_bias)
     projection = paddle.reshape(projection, shape=[-1, vocab_size])
 
     loss = layers.softmax_with_cross_entropy(

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

@@ -157,7 +157,7 @@ class EagerDeletionRecurrentOpTest1(unittest.TestCase):
             x_t = rnn.step_input(x)
 
             h = paddle.scale(
-                x=layers.elementwise_add(x=h_pre, y=x_t),
+                x=paddle.add(x=h_pre, y=x_t),
                 scale=self.py_rnn.scale,
             )
 
@@ -328,9 +328,7 @@ class EagerDeletionRecurrentOpTest2(EagerDeletionRecurrentOpTest1):
                 bias_attr=False,
             )
 
-            h = paddle.nn.functional.sigmoid(
-                x=layers.elementwise_add(x=temp_l, y=temp_r)
-            )
+            h = paddle.nn.functional.sigmoid(x=paddle.add(x=temp_l, y=temp_r))
 
             rnn.update_memory(h_pre, h)
             rnn.output(h)
@@ -504,7 +502,7 @@ class EagerDeletionRecurrentOpNoMemBootTest(EagerDeletionRecurrentOpTest1):
         with rnn.step():
             mem_pre = rnn.memory(shape=[-1, self.input_dim], batch_ref=x)
             x_t = rnn.step_input(x)
-            mem = layers.elementwise_add(x=mem_pre, y=x_t)
+            mem = paddle.add(x=mem_pre, y=x_t)
             rnn.update_memory(mem_pre, mem)
             rnn.output(mem)
 
@@ -584,7 +582,7 @@ class EagerDeletionTwoRecurrentOpsTest(EagerDeletionRecurrentOpTest1):
         with rnn_0.step():
             x_t = rnn_0.step_input(x)
             mem_pre = rnn_0.memory(shape=[-1, self.input_dim], batch_ref=x)
-            mem = layers.elementwise_add(x=mem_pre, y=x_t)
+            mem = paddle.add(x=mem_pre, y=x_t)
             rnn_0.update_memory(mem_pre, mem)
             rnn_0.output(mem)
 
@@ -594,8 +592,8 @@ class EagerDeletionTwoRecurrentOpsTest(EagerDeletionRecurrentOpTest1):
             x_t = rnn_1.step_input(x)
             last_rnn_output = rnn_0()
             last_rnn_sum = paddle.sum(last_rnn_output)
-            mem = layers.elementwise_add(x=x_t, y=last_rnn_sum)
-            y = layers.elementwise_add(x=mem_pre, y=mem)
+            mem = paddle.add(x=x_t, y=last_rnn_sum)
+            y = paddle.add(x=mem_pre, y=mem)
             rnn_1.update_memory(mem_pre, mem)
             rnn_1.output(y)
         return rnn_1()
@@ -693,7 +691,7 @@ class EagerDeletionFarwardOnlyRnnAndBackwardRnnTest(
                 x_t = forward_only_rnn.step_input(x)
 
                 h = paddle.scale(
-                    x=layers.elementwise_add(x=h_pre, y=x_t),
+                    x=paddle.add(x=h_pre, y=x_t),
                     scale=self.py_rnn.scale,
                 )
 
@@ -709,7 +707,7 @@ class EagerDeletionFarwardOnlyRnnAndBackwardRnnTest(
                 x_t = rnn.step_input(x)
 
                 h = paddle.scale(
-                    x=layers.elementwise_add(x=h_pre, y=x_t),
+                    x=paddle.add(x=h_pre, y=x_t),
                     scale=self.py_rnn.scale,
                 )
 

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

@@ -94,7 +94,7 @@ class TestElementwiseAddDoubleGradCheck(unittest.TestCase):
         y = layers.data('y', shape, False, dtype)
         x.persistable = True
         y.persistable = True
-        out = layers.elementwise_add(x, y)
+        out = paddle.add(x, y)
         x_arr = np.random.uniform(-1, 1, shape).astype(dtype)
         y_arr = np.random.uniform(-1, 1, shape).astype(dtype)
 
@@ -155,7 +155,7 @@ class TestElementwiseSubDoubleGradCheck(unittest.TestCase):
         y = layers.data('y', shape, False, dtype)
         x.persistable = True
         y.persistable = True
-        out = layers.elementwise_sub(x, y)
+        out = paddle.subtract(x, y)
         x_arr = np.random.uniform(-1, 1, shape).astype(dtype)
         y_arr = np.random.uniform(-1, 1, shape).astype(dtype)
 
@@ -291,7 +291,7 @@ class TestElementwiseAddTripleGradCheck(unittest.TestCase):
         y = layers.data('y', shape, False, dtype)
         x.persistable = True
         y.persistable = True
-        out = layers.elementwise_add(x, y)
+        out = paddle.add(x, y)
         x_arr = np.random.uniform(-1, 1, shape).astype(dtype)
         y_arr = np.random.uniform(-1, 1, shape).astype(dtype)
 

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

@@ -14,6 +14,7 @@
 
 import unittest
 
+import paddle
 import paddle.fluid as fluid
 import paddle.fluid.nets as nets
 from paddle.fluid.framework import Program
@@ -81,7 +82,7 @@ class TestLayer(unittest.TestCase):
             image2 = fluid.layers.data(
                 name='pixel2', shape=[3, 48, 48], dtype='float32'
             )
-            fluid.layers.elementwise_add(x=image1, y=image2, act='relu')
+            paddle.nn.functional.relu(paddle.add(x=image1, y=image2))
         print(main_program)
 
 

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

@@ -33,7 +33,7 @@ class MyLayer(fluid.Layer):
     def forward(self, inputs):
         x = fluid.layers.relu(inputs)
         self._x_for_debug = x
-        x = fluid.layers.elementwise_mul(x, x)
+        x = paddle.multiply(x, x)
         x = paddle.sum(x)
         return [x]
 
@@ -722,9 +722,9 @@ class TestImperative(unittest.TestCase):
             inp1 = paddle.to_tensor(np_inp1)
             inp2 = paddle.to_tensor(np_inp2)
             if np.sum(np_inp1) < np.sum(np_inp2):
-                x = fluid.layers.elementwise_add(inp1, inp2)
+                x = paddle.add(inp1, inp2)
             else:
-                x = fluid.layers.elementwise_sub(inp1, inp2)
+                x = paddle.subtract(inp1, inp2)
             dygraph_result = x.numpy()
 
         # static graph
@@ -750,13 +750,13 @@ class TestImperative(unittest.TestCase):
             with ie.true_block():
                 d1 = ie.input(inp_data1)
                 d2 = ie.input(inp_data2)
-                d3 = fluid.layers.elementwise_add(d1, d2)
+                d3 = paddle.add(d1, d2)
                 ie.output(d3)
 
             with ie.false_block():
                 d1 = ie.input(inp_data1)
                 d2 = ie.input(inp_data2)
-                d3 = fluid.layers.elementwise_sub(d1, d2)
+                d3 = paddle.subtract(d1, d2)
                 ie.output(d3)
             out = ie()
 

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

@@ -76,7 +76,7 @@ class DMF(fluid.Layer):
         for ul, il in zip(self._user_layers, self._item_layers):
             users = ul(users)
             items = il(items)
-        return fluid.layers.elementwise_mul(users, items)
+        return paddle.multiply(users, items)
 
 
 class MLP(fluid.Layer):

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

@@ -67,7 +67,7 @@ class SimpleNet(fluid.Layer):
         projection = fluid.layers.matmul(
             x_emb, paddle.transpose(self.embedding.weight, perm=[1, 0])
         )
-        projection = fluid.layers.elementwise_add(projection, self.softmax_bias)
+        projection = paddle.add(projection, self.softmax_bias)
         projection = paddle.reshape(projection, shape=[-1, self.vocab_size])
         loss = fluid.layers.softmax_with_cross_entropy(
             logits=projection, label=label, soft_label=False

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

@@ -306,9 +306,7 @@ class SimpleAttention(fluid.dygraph.Layer):
             decoder_state_proj_reshape,
             [-1, encoder_proj.shape[1], -1],
         )
-        concated = fluid.layers.elementwise_add(
-            encoder_proj, decoder_state_expand
-        )
+        concated = paddle.add(encoder_proj, decoder_state_expand)
         concated = paddle.tanh(x=concated)
         attention_weight = self.fc_2(concated)
 
@@ -362,7 +360,7 @@ class GRUDecoderWithAttention(fluid.dygraph.Layer):
             )
             fc_1 = self.fc_1_layer(context)
             fc_2 = self.fc_2_layer(current_word)
-            decoder_inputs = fluid.layers.elementwise_add(x=fc_1, y=fc_2)
+            decoder_inputs = paddle.add(x=fc_1, y=fc_2)
 
             h, _, _ = self.gru_unit(decoder_inputs, hidden_mem)
             hidden_mem = h

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

@@ -35,7 +35,7 @@ class MyLayer(fluid.Layer):
 
     def forward(self, inputs):
         x = fluid.layers.relu(inputs)
-        x = fluid.layers.elementwise_mul(x, x)
+        x = paddle.multiply(x, x)
         x = paddle.sum(x)
         return [x]