!19081 add ops IsNan

Merge pull request !19081 from lingyunli63/add_ops_IsNan

mindspore/_extends/graph_kernel/expanders/__init__.py

@@ -33,6 +33,7 @@ from .fused_mul_add import FusedMulAdd
 from .gelu import GeLU
 from .gelu_grad import GeLUGrad
 from .gkdropout import GkDropout
+from .identity import Identity
 from .lamb_apply_optimizer_assign import LambApplyOptimizerAssign
 from .lamb_apply_weight_assign import LambApplyWeightAssign
 from .layernorm import LayerNorm
@@ -42,6 +43,7 @@ from .logsoftmax_grad import LogSoftmaxGrad
 from .matmul import BatchMatMul, MatMul
 from .maximum_grad import MaximumGrad
 from .minimum_grad import MinimumGrad
+from .oneslike import OnesLike
 from .reduce_mean import ReduceMean
 from .relu import ReLU
 from .relu_grad import ReluGrad
@@ -56,6 +58,7 @@ from .sqrt_grad import SqrtGrad
 from .square import Square
 from .square_sum_v1 import SquareSumV1
 from .squared_difference import SquaredDifference
+from .square_sum_all import SquareSumAll
 from .squeeze import Squeeze
 from .tanh_grad import TanhGrad
 from .tile import Tile

mindspore/_extends/graph_kernel/expanders/identity.py

@@ -0,0 +1,23 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ===========================================================================
+"""generate json desc for Identity"""
+from ._utils import Expander
+
+class Identity(Expander):
+    """Identity expander"""
+    def _expand(self, graph_builder):
+        input_x = self.inputs[0]
+        result = graph_builder.emit('Reshape', [input_x], attrs={'shape': input_x.shape})
+        return result

mindspore/_extends/graph_kernel/expanders/oneslike.py

@@ -0,0 +1,24 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ===========================================================================
+"""generate json desc for OnesLike"""
+from ._utils import Expander
+
+class OnesLike(Expander):
+    """OnesLike expander"""
+    def _expand(self, graph_builder):
+        input_x = self.inputs[0]
+        const_one = graph_builder.value(input_x.dtype, 1)
+        result = graph_builder.emit('BroadcastTo', [const_one], attrs={'shape': input_x.shape})
+        return result

mindspore/_extends/graph_kernel/expanders/square_sum_all.py

@@ -0,0 +1,41 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ===========================================================================
+"""generate json desc for SquareSumAll"""
+from ._utils import Expander
+
+class SquareSumAll(Expander):
+    """SquareSumAll expander"""
+    def _check(self):
+        """check inputs"""
+        input_num = len(self.inputs)
+        if input_num != 2:
+            raise GKException("SquareSumAll inputs number should be 2, but got {}.".format(input_num))
+
+    def _expand(self, graph_builder):
+        """do expand"""
+        x0 = self.inputs[0]
+        x1 = self.inputs[1]
+
+        ori_shape = x0.shape
+        axis = []
+        for i, _ in enumerate(ori_shape):
+            axis.append(i)
+
+        square_res0 = graph_builder.emit('Mul', [x0, x0])
+        square_res1 = graph_builder.emit('Mul', [x1, x1])
+        result0 = graph_builder.emit('ReduceSum', [square_res0], attrs={'reduce_axis': axis, 'keep_dims': False})
+        result1 = graph_builder.emit('ReduceSum', [square_res1], attrs={'reduce_axis': axis, 'keep_dims': False})
+
+        return result0, result1

mindspore/_extends/graph_kernel/model/model.py

@@ -155,6 +155,9 @@ class PrimLib:
         'Mul': Prim(ELEMWISE),
         'Sub': Prim(ELEMWISE),
         'Log': Prim(ELEMWISE),
+        'IsNan': Prim(ELEMWISE),
+        'IsInf': Prim(ELEMWISE),
+        'IsFinite': Prim(ELEMWISE),
         'Exp': Prim(ELEMWISE),
         'Rsqrt': Prim(ELEMWISE),
         'Sqrt': Prim(ELEMWISE),

mindspore/ccsrc/backend/optimizer/graph_kernel/graph_kernel_cluster.cc

@@ -94,6 +94,9 @@ std::vector<PrimitivePtr> GetClusterableOpList() {
     prim::kPrimLogicalAnd,
     prim::kPrimLogicalOr,
     prim::kPrimLogicalNot,
+    prim::kPrimIsNan,
+    prim::kPrimIsInf,
+    prim::kPrimIsFinite,
 #endif
   };
   const auto &flags = context::GraphKernelFlags::GetInstance();

mindspore/ccsrc/backend/optimizer/graph_kernel/graph_kernel_expander.cc

@@ -88,6 +88,9 @@ std::vector<PrimitivePtr> GetExpandOps() {
     prim::kPrimSquaredDifference,
     prim::kPrimSqueeze,
     prim::kPrimEqualCount,
+    prim::kPrimSquareSumAll,
+    prim::kPrimIdentityMath,
+    prim::kPrimOnesLike,
 #endif
   };
   const auto &flags = context::GraphKernelFlags::GetInstance();

mindspore/ccsrc/runtime/framework/graph_scheduler.cc

@@ -2580,7 +2580,9 @@ bool GraphScheduler::CheckActorValid(const ActorSet *actor_set, GraphExecutionSt
   // Check the data source actors.
   for (const auto &data_source_actor : actor_set->data_source_actors_) {
     MS_EXCEPTION_IF_NULL(data_source_actor);
-    if (data_source_actor->output_data_arrows_.size() + data_source_actor->output_result_arrows_.size() == 0) {
+    if (data_source_actor->output_data_arrows_.size() + data_source_actor->output_result_arrows_.size() +
+          data_source_actor->output_control_arrows_.size() ==
+        0) {
       MS_LOG(ERROR) << data_source_actor->GetAID().Name() << " has no user.";
       return false;
     }

mindspore/core/base/core_ops.h

@@ -488,6 +488,10 @@ inline const PrimitivePtr kPrimAtanGrad = std::make_shared<Primitive>("AtanGrad"
 inline const PrimitivePtr kPrimFloorMod = std::make_shared<Primitive>("FloorMod");
 inline const PrimitivePtr kPrimWhere = std::make_shared<Primitive>("Where");
 inline const PrimitivePtr kPrimIdentityMath = std::make_shared<Primitive>("Identity", kSideEffectPropagate);
+inline const PrimitivePtr kPrimIsNan = std::make_shared<Primitive>("IsNan");
+inline const PrimitivePtr kPrimIsInf = std::make_shared<Primitive>("IsInf");
+inline const PrimitivePtr kPrimIsFinite = std::make_shared<Primitive>("IsFinite");
+inline const PrimitivePtr kPrimSquareSumAll = std::make_shared<Primitive>("SquareSumAll");
 
 // Statements
 inline const PrimitivePtr kPrimReturn = std::make_shared<Primitive>("Return");

tests/st/ops/graph_kernel/test_identity.py

@@ -0,0 +1,56 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+
+import numpy as np
+import pytest
+import mindspore.context as context
+from mindspore import Tensor
+from mindspore.nn import Cell
+import mindspore.ops.operations as P
+
+class Net(Cell):
+    def __init__(self):
+        super(Net, self).__init__()
+        self.identity = P.Identity()
+
+    def construct(self, x):
+        return self.identity(x)
+
+def get_output(x, enable_graph_kernel=False):
+    context.set_context(enable_graph_kernel=enable_graph_kernel)
+    net = Net()
+    output = net(x)
+    return output
+
+def test_basic(dtype):
+    expect_np = np.random.normal(0, 10, (16, 32)).astype(dtype)
+    x = Tensor(expect_np)
+    output = get_output(x, True)
+    output_np = output.asnumpy().copy()
+    assert np.allclose(expect_np, output_np, 1.e-4, 1.e-7)
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gpu_1():
+    context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
+    test_basic(np.float16)
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gpu_2():
+    context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
+    test_basic(np.float32)

tests/st/ops/graph_kernel/test_oneslike.py

@@ -0,0 +1,56 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+
+import numpy as np
+import pytest
+import mindspore.context as context
+from mindspore import Tensor
+from mindspore.nn import Cell
+import mindspore.ops.operations as P
+import mindspore.common.dtype as mstype
+
+class Net(Cell):
+    def __init__(self):
+        super(Net, self).__init__()
+        self.oneslike = P.OnesLike()
+
+    def construct(self, shape, dtype, x):
+        return self.oneslike(x)
+
+def get_output(shape, dtype, nptype, enable_graph_kernel=False):
+    context.set_context(enable_graph_kernel=enable_graph_kernel)
+    net = Net()
+    x = Tensor(np.random.normal(0, 10, shape).astype(nptype))
+    output = net(shape, dtype, x)
+    return output
+
+def test_basic(shape, dtype, nptype):
+    expect = get_output(shape, dtype, nptype, False)
+    output = get_output(shape, dtype, nptype, True)
+    assert np.allclose(expect.asnumpy(), output.asnumpy(), 1.e-4, 1.e-7)
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gpu_1():
+    context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
+    test_basic((2, 16), mstype.float16, np.float16)
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gpu_2():
+    context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
+    test_basic((4, 32), mstype.float32, np.float32)

tests/st/ops/graph_kernel/test_square_sum_all.py

@@ -0,0 +1,61 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+
+import numpy as np
+import pytest
+import mindspore.context as context
+from mindspore import Tensor
+from mindspore.nn import Cell
+import mindspore.ops.operations as P
+
+class Net(Cell):
+    def __init__(self):
+        super(Net, self).__init__()
+        self.squaresumall = P.SquareSumAll()
+
+    def construct(self, x0, x1):
+        return self.squaresumall(x0, x1)
+
+def get_output(inp0, inp1, enable_graph_kernel=False):
+    context.set_context(enable_graph_kernel=enable_graph_kernel)
+    net = Net()
+    output = net(inp0, inp1)
+    return output
+
+def test_basic(datatype):
+    inp0 = Tensor(np.random.normal(1, 0.1, [800, 96]).astype(datatype))
+    inp1 = Tensor(np.random.normal(1, 0.1, [800, 96]).astype(datatype))
+    expect = get_output(inp0, inp1, False)
+    output = get_output(inp0, inp1, True)
+    expect_np0 = expect[0].asnumpy().copy()
+    output_np0 = output[0].asnumpy().copy()
+    expect_np1 = expect[1].asnumpy().copy()
+    output_np1 = output[1].asnumpy().copy()
+    assert np.allclose(expect_np0, output_np0, 1.e-4, 1.e-7)
+    assert np.allclose(expect_np1, output_np1, 1.e-4, 1.e-7)
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gpu_1():
+    context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
+    test_basic(np.float16)
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gpu_2():
+    context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
+    test_basic(np.float32)