rm unittests eager guard tests part9 histogram2imperative_dataloader (#…

…48825)

* rm unittests eager guard tests part9 histogram2imperative_dataloader

* rm basic

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

@@ -20,7 +20,6 @@
 import paddle
 import paddle.fluid as fluid
 from paddle.fluid import Program, program_guard
-from paddle.fluid.framework import _test_eager_guard
 
 
 class TestHistogramOpAPI(unittest.TestCase):
@@ -59,15 +58,13 @@ def test_dygraph(self):
                 msg='histogram output is wrong, out =' + str(actual.numpy()),
             )
 
-            with _test_eager_guard():
-                inputs_np = np.array([[2, 4, 2], [2, 5, 4]]).astype(np.int64)
-                inputs = paddle.to_tensor(inputs_np)
-                actual = paddle.histogram(inputs, bins=5, min=1, max=5)
-                self.assertTrue(
-                    (actual.numpy() == expected).all(),
-                    msg='histogram output is wrong, out ='
-                    + str(actual.numpy()),
-                )
+            inputs_np = np.array([[2, 4, 2], [2, 5, 4]]).astype(np.int64)
+            inputs = paddle.to_tensor(inputs_np)
+            actual = paddle.histogram(inputs, bins=5, min=1, max=5)
+            self.assertTrue(
+                (actual.numpy() == expected).all(),
+                msg='histogram output is wrong, out =' + str(actual.numpy()),
+            )
 
 
 class TestHistogramOpError(unittest.TestCase):

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

@@ -18,7 +18,6 @@
 
 import paddle
 import paddle.fluid as fluid
-from paddle.fluid.framework import _test_eager_guard
 from paddle.nn import Embedding
 from paddle.tensor import random
 
@@ -169,7 +168,7 @@ def embed_linear0(self, x):
 
 
 class TestImperativeAutoPrune(unittest.TestCase):
-    def func_auto_prune(self):
+    def test_auto_prune(self):
         with fluid.dygraph.guard():
             case1 = AutoPruneLayer0(input_size=5)
             value1 = np.arange(25).reshape(5, 5).astype("float32")
@@ -181,12 +180,7 @@ def func_auto_prune(self):
             self.assertIsNotNone(case1.linear2.weight._grad_ivar())
             self.assertIsNotNone(case1.linear1.weight._grad_ivar())
 
-    def test_auto_prune(self):
-        with _test_eager_guard():
-            self.func_auto_prune()
-        self.func_auto_prune()
-
-    def func_auto_prune2(self):
+    def test_auto_prune2(self):
         with fluid.dygraph.guard():
             case2 = AutoPruneLayer1(input_size=5)
             value1 = np.arange(25).reshape(5, 5).astype("float32")
@@ -199,13 +193,9 @@ def func_auto_prune2(self):
             self.assertIsNone(case2.linear2.weight._grad_ivar())
             self.assertIsNotNone(case2.linear1.weight._grad_ivar())
 
-    def test_auto_prune2(self):
-        with _test_eager_guard():
-            self.func_auto_prune2()
-        self.func_auto_prune2()
-
     # TODO(jiabin): Support this when we support better split tensor
-    def func_auto_prune3(self):
+    def test_auto_prune3(self):
+        fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": True})
         with fluid.dygraph.guard():
             case3 = AutoPruneLayer3(input_size=784)
             value1 = np.arange(784).reshape(1, 784).astype("float32")
@@ -216,15 +206,10 @@ def func_auto_prune3(self):
             loss.backward()
             self.assertIsNotNone(case3.linear.weight._grad_ivar())
             self.assertTrue((part2.gradient() == 0).all())
-
-    def test_auto_prune3(self):
-        fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": True})
-        with _test_eager_guard():
-            self.func_auto_prune3()
-        self.func_auto_prune3()
         fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": False})
 
-    def func_auto_prune4(self):
+    def test_auto_prune4(self):
+        fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": True})
         with fluid.dygraph.guard():
             case4 = AutoPruneLayer3(input_size=784)
             value1 = np.arange(784).reshape(1, 784).astype("float32")
@@ -235,15 +220,10 @@ def func_auto_prune4(self):
             part2.backward()
             self.assertIsNotNone(case4.linear.weight._grad_ivar())
             self.assertTrue((part2.gradient() == 1).all())
-
-    def test_auto_prune4(self):
-        fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": True})
-        with _test_eager_guard():
-            self.func_auto_prune4()
-        self.func_auto_prune4()
         fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": False})
 
-    def func_auto_prune5(self):
+    def test_auto_prune5(self):
+        fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": True})
         with fluid.dygraph.guard():
             case4 = AutoPruneLayer3(input_size=784)
             value1 = np.arange(784).reshape(1, 784).astype("float32")
@@ -254,15 +234,9 @@ def func_auto_prune5(self):
             part1.backward()
             self.assertIsNotNone(case4.linear.weight._grad_ivar())
             self.assertTrue((part2.gradient() == 0).all())
-
-    def test_auto_prune5(self):
-        fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": True})
-        with _test_eager_guard():
-            self.func_auto_prune5()
-        self.func_auto_prune5()
         fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": False})
 
-    def func_auto_prune6(self):
+    def test_auto_prune6(self):
         with fluid.dygraph.guard():
             value0 = np.arange(26).reshape(2, 13).astype("float32")
             value1 = np.arange(6).reshape(2, 3).astype("float32")
@@ -280,12 +254,7 @@ def func_auto_prune6(self):
             self.assertIsNone(linear.weight.gradient())
             self.assertIsNone(out1.gradient())
 
-    def test_auto_prune6(self):
-        with _test_eager_guard():
-            self.func_auto_prune6()
-        self.func_auto_prune6()
-
-    def func_auto_prune7(self):
+    def test_auto_prune7(self):
         with fluid.dygraph.guard():
             value0 = np.arange(26).reshape(2, 13).astype("float32")
             value1 = np.arange(6).reshape(2, 3).astype("float32")
@@ -303,12 +272,7 @@ def func_auto_prune7(self):
             self.assertIsNone(linear.weight.gradient())
             self.assertIsNone(out1.gradient())
 
-    def test_auto_prune7(self):
-        with _test_eager_guard():
-            self.func_auto_prune7()
-        self.func_auto_prune7()
-
-    def func_auto_prune8(self):
+    def test_auto_prune8(self):
         with fluid.dygraph.guard():
             value0 = np.arange(26).reshape(2, 13).astype("float32")
             value1 = np.arange(6).reshape(2, 3).astype("float32")
@@ -336,12 +300,7 @@ def func_auto_prune8(self):
                 np.array_equal(linear_origin, linear.weight.numpy())
             )
 
-    def test_auto_prune8(self):
-        with _test_eager_guard():
-            self.func_auto_prune8()
-        self.func_auto_prune8()
-
-    def func_auto_prune9(self):
+    def test_auto_prune9(self):
         with fluid.dygraph.guard():
             value0 = np.arange(26).reshape(2, 13).astype("float32")
             value1 = np.arange(6).reshape(2, 3).astype("float32")
@@ -371,12 +330,7 @@ def func_auto_prune9(self):
             except ValueError as e:
                 assert type(e) == ValueError
 
-    def test_auto_prune9(self):
-        with _test_eager_guard():
-            self.func_auto_prune9()
-        self.func_auto_prune9()
-
-    def func_auto_prune10(self):
+    def test_auto_prune10(self):
         with fluid.dygraph.guard():
             value0 = np.arange(26).reshape(2, 13).astype("float32")
             value1 = np.arange(6).reshape(2, 3).astype("float32")
@@ -396,12 +350,7 @@ def func_auto_prune10(self):
             self.assertIsNone(linear.weight.gradient())
             self.assertIsNone(out1.gradient())
 
-    def test_auto_prune10(self):
-        with _test_eager_guard():
-            self.func_auto_prune10()
-        self.func_auto_prune10()
-
-    def func_auto_prune_with_optimizer(self):
+    def test_auto_prune_with_optimizer(self):
         vocab_size = 100
         size = 20
         batch_size = 16
@@ -451,12 +400,7 @@ def func_auto_prune_with_optimizer(self):
             assert model.embed1.weight._grad_ivar() is None
             assert model.linear_1.weight._grad_ivar() is None
 
-    def test_auto_prune_with_optimizer(self):
-        with _test_eager_guard():
-            self.func_auto_prune_with_optimizer()
-        self.func_auto_prune_with_optimizer()
-
-    def func_case2_prune_no_grad_branch(self):
+    def test_case2_prune_no_grad_branch(self):
         with fluid.dygraph.guard():
             value1 = np.arange(784).reshape(1, 784)
             value2 = np.arange(1).reshape(1, 1)
@@ -468,12 +412,7 @@ def func_case2_prune_no_grad_branch(self):
             self.assertIsNone(case3.linear2.weight._grad_ivar())
             self.assertIsNotNone(case3.linear.weight._grad_ivar())
 
-    def test_case2_prune_no_grad_branch(self):
-        with _test_eager_guard():
-            self.func_case2_prune_no_grad_branch()
-        self.func_case2_prune_no_grad_branch()
-
-    def func_case3_prune_no_grad_branch2(self):
+    def test_case3_prune_no_grad_branch2(self):
         with fluid.dygraph.guard():
             value1 = np.arange(1).reshape(1, 1)
             linear = paddle.nn.Linear(1, 1)
@@ -486,23 +425,13 @@ def func_case3_prune_no_grad_branch2(self):
             loss.backward()
             self.assertIsNone(linear.weight._grad_ivar())
 
-    def test_case3_prune_no_grad_branch2(self):
-        with _test_eager_guard():
-            self.func_case3_prune_no_grad_branch2()
-        self.func_case3_prune_no_grad_branch2()
-
-    def func_case4_with_no_grad_op_maker(self):
+    def test_case4_with_no_grad_op_maker(self):
         with fluid.dygraph.guard():
             out = random.gaussian(shape=[20, 30])
             loss = paddle.mean(out)
             loss.backward()
             self.assertIsNone(out._grad_ivar())
 
-    def test_case4_with_no_grad_op_maker(self):
-        with _test_eager_guard():
-            self.func_case4_with_no_grad_op_maker()
-        self.func_case4_with_no_grad_op_maker()
-
 
 if __name__ == '__main__':
     unittest.main()

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

@@ -16,11 +16,10 @@
 from collections import OrderedDict
 
 import paddle
-from paddle.fluid.framework import _test_eager_guard
 
 
 class TestLayerDict(unittest.TestCase):
-    def func_layer_dict(self):
+    def test_layer_dict(self):
         layers = OrderedDict(
             [
                 ('conv1d', paddle.nn.Conv1D(3, 2, 3)),
@@ -92,12 +91,7 @@ def check_layer_dict():
         layers_dicts.update(list_format_layers)
         check_layer_dict()
 
-    def test_layer_dict(self):
-        with _test_eager_guard():
-            self.func_layer_dict()
-        self.func_layer_dict()
-
-    def func_layer_dict_error_inputs(self):
+    def test_layer_dict_error_inputs(self):
         layers = [
             ('conv1d', paddle.nn.Conv1D(3, 2, 3), "conv1d"),
             ('conv2d', paddle.nn.Conv2D(3, 2, 3)),
@@ -108,11 +102,6 @@ def func_layer_dict_error_inputs(self):
 
         self.assertRaises(AssertionError, layers_dicts.update, 1)
 
-    def test_layer_dict_error_inputs(self):
-        with _test_eager_guard():
-            self.func_layer_dict_error_inputs()
-        self.func_layer_dict_error_inputs()
-
 
 if __name__ == '__main__':
     unittest.main()

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

@@ -18,7 +18,6 @@
 
 import paddle
 import paddle.fluid as fluid
-from paddle.fluid.framework import _test_eager_guard
 
 
 class MyLayer(fluid.Layer):
@@ -94,15 +93,10 @@ def layer_list(self, use_fluid_api):
             self.assertListEqual(res11.shape, [5, 4])
             res11.backward()
 
-    def func_test_layer_list(self):
+    def test_test_layer_list(self):
         self.layer_list(True)
         self.layer_list(False)
 
-    def test_layer_list(self):
-        with _test_eager_guard():
-            self.func_test_layer_list()
-        self.func_test_layer_list()
-
 
 if __name__ == '__main__':
     unittest.main()

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

@@ -19,7 +19,6 @@
 import paddle
 import paddle.fluid as fluid
 from paddle import _legacy_C_ops
-from paddle.fluid.framework import _test_eager_guard
 
 
 class MyLayer(fluid.Layer):
@@ -68,9 +67,6 @@ def paramter_list(self, use_fluid_api):
             loss.backward()
 
     def test_paramter_list(self):
-        with _test_eager_guard():
-            self.paramter_list(False)
-            self.paramter_list(True)
         self.paramter_list(False)
         self.paramter_list(True)
 

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

@@ -18,12 +18,11 @@
 
 import paddle
 import paddle.fluid as fluid
-from paddle.fluid.framework import _test_eager_guard
 from paddle.nn import Linear
 
 
 class TestImperativeContainerSequential(unittest.TestCase):
-    def func_sequential(self):
+    def test_sequential(self):
         data = np.random.uniform(-1, 1, [5, 10]).astype('float32')
         with fluid.dygraph.guard():
             data = fluid.dygraph.to_variable(data)
@@ -57,12 +56,7 @@ def func_sequential(self):
             loss2 = paddle.mean(res2)
             loss2.backward()
 
-    def test_sequential(self):
-        with _test_eager_guard():
-            self.func_sequential()
-        self.func_sequential()
-
-    def func_sequential_list_params(self):
+    def test_sequential_list_params(self):
         data = np.random.uniform(-1, 1, [5, 10]).astype('float32')
         with fluid.dygraph.guard():
             data = fluid.dygraph.to_variable(data)
@@ -96,11 +90,6 @@ def func_sequential_list_params(self):
             loss2 = paddle.mean(res2)
             loss2.backward()
 
-    def test_sequential_list_params(self):
-        with _test_eager_guard():
-            self.func_sequential_list_params()
-        self.func_sequential_list_params()
-
 
 if __name__ == '__main__':
     unittest.main()