Nigel
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

!14805 [ME][ControlFlow] Add some control sink testcases 

From: @Margaret_wangrui
Reviewed-by: @ginfung,@zh_qh
Signed-off-by: @zh_qh
This commit is contained in:
mindspore-ci-bot 2021-04-09 09:18:24 +08:00 committed by Gitee
parent bdc5a9c88b e69ead3f19
commit 32f92eecbd
10 changed files with 798 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

65
tests/st/control/inner/test_002_single_for.py Normal file
View File

@ -0,0 +1,65 @@
# 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.
# ============================================================================
from mindspore import context
from mindspore import Tensor, nn
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_signle_for():
class SignleForNet(nn.Cell):
def __init__(self):
super().__init__()
self.add = P.Add()
self.mul = P.Mul()
def construct(self, x, y, z):
x = self.add(x, y)
for _ in range(0, 3):
z = self.add(z, x)
y = self.mul(z, y)
return y
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([2], mstype.int32)
y = Tensor([5], mstype.int32)
z = Tensor([4], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_net = SignleForNet()
net = GradNet(for_net)
graph_forward_res = for_net(x, y, z)
graph_backward_res = net(x, y, z)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_net = SignleForNet()
net = GradNet(for_net)
pynative_forward_res = for_net(x, y, z)
pynative_backward_res = net(x, y, z)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

70
tests/st/control/inner/test_030_for_in_if.py Normal file
View File

@ -0,0 +1,70 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_for_in_if():
class ForInIfNet(nn.Cell):
def __init__(self):
super().__init__()
self.mul = P.Mul()
self.add = P.Add()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 4, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
def construct(self, x):
if self.param_a > self.param_b:
x = self.mul(x, 2)
for _ in range(0, 5):
x = self.add(x, x)
self.param_b += 1
return x
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([10], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_in_if_net = ForInIfNet()
net = GradNet(for_in_if_net)
graph_forward_res = for_in_if_net(x)
graph_backward_res = net(x)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_in_if_net = ForInIfNet()
net = GradNet(for_in_if_net)
pynative_forward_res = for_in_if_net(x)
pynative_backward_res = net(x)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

75
tests/st/control/inner/test_031_for_in_while.py Normal file
View File

@ -0,0 +1,75 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_for_in_while():
class ForInWhileNet(nn.Cell):
def __init__(self):
super().__init__()
self.mul = P.Mul()
self.add = P.Add()
self.sub = P.Sub()
self.assign = P.Assign()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 2, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
def construct(self, x):
self.assign(self.param_a, x + self.param_a)
while self.param_a > self.param_b:
x = self.mul(x, 2)
for _ in range(0, 5):
x = self.add(x, x)
self.param_b = self.param_b + 1
y = self.sub(x, self.param_b)
self.assign(self.param_a, y)
return x
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([2], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_in_while_net = ForInWhileNet()
net = GradNet(for_in_while_net)
graph_forward_res = for_in_while_net(x)
graph_backward_res = net(x)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_in_while_net = ForInWhileNet()
net = GradNet(for_in_while_net)
pynative_forward_res = for_in_while_net(x)
pynative_backward_res = net(x)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

75
tests/st/control/inner/test_032_for_in_for.py Normal file
View File

@ -0,0 +1,75 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_for_in_for():
class ForInForNet(nn.Cell):
def __init__(self):
super().__init__()
self.relu = nn.ReLU()
self.mul = P.Mul()
self.add = P.Add()
self.sub = P.Sub()
self.assign = P.Assign()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 2, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
def construct(self, x):
self.assign(self.param_a, x + self.param_a)
for _ in range(0, 10):
x = self.mul(x, 2)
for _ in range(0, 5):
x = self.add(x, x)
self.param_b += 1
y = self.sub(x, self.param_b)
z = self.relu(x + y)
return z
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([2], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_in_for_net = ForInForNet()
net = GradNet(for_in_for_net)
graph_forward_res = for_in_for_net(x)
graph_backward_res = net(x)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_in_for_net = ForInForNet()
net = GradNet(for_in_for_net)
pynative_forward_res = for_in_for_net(x)
pynative_backward_res = net(x)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

77
tests/st/control/inner/test_102_if_after_for.py Normal file
View File

@ -0,0 +1,77 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_if_after_for():
class IfAfterForNet(nn.Cell):
def __init__(self):
super().__init__()
self.relu = nn.ReLU()
self.mul = P.Mul()
self.add = P.Add()
self.sub = P.Sub()
self.assign = P.Assign()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 2, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
def construct(self, x):
self.assign(self.param_a, x + self.param_a)
y = self.add(x, self.param_b)
for _ in range(0, 2):
x = self.sub(x, 2)
self.param_b = self.add(self.param_b, 2)
if x < self.param_b:
y = self.mul(x, self.param_a)
z = self.relu(x + y)
return z
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([7], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
if_after_for_net = IfAfterForNet()
net = GradNet(if_after_for_net)
graph_forward_res = if_after_for_net(x)
graph_backward_res = net(x)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
if_after_for_net = IfAfterForNet()
net = GradNet(if_after_for_net)
pynative_forward_res = if_after_for_net(x)
pynative_backward_res = net(x)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

79
tests/st/control/inner/test_302_for_after_for.py Normal file
View File

@ -0,0 +1,79 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_for_after_for():
class ForAfterForNet(nn.Cell):
def __init__(self):
super().__init__()
self.relu = nn.ReLU()
self.mul = P.Mul()
self.add = P.Add()
self.sub = P.Sub()
self.div = P.Div()
self.assign = P.Assign()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 2, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
def construct(self, x):
self.assign(self.param_a, x + self.param_a)
y = self.add(x, self.param_b)
for _ in range(0, 2):
x = self.sub(x, 3)
y = x + self.param_b
self.param_a = x + y
for _ in range(0, 5):
y = self.mul(x, self.param_a)
z = self.relu(y + self.param_a)
return z
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([7], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_after_for_net = ForAfterForNet()
net = GradNet(for_after_for_net)
graph_forward_res = for_after_for_net(x)
graph_backward_res = net(x)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_after_for_net = ForAfterForNet()
net = GradNet(for_after_for_net)
pynative_forward_res = for_after_for_net(x)
pynative_backward_res = net(x)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

92
tests/st/control/inner/test_320_for_after_while_in_if.py Normal file
View File

@ -0,0 +1,92 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_for_after_while_in_if():
class ForAfterWhileInIfNet(nn.Cell):
def __init__(self):
super().__init__()
self.relu = nn.ReLU()
self.softmax = nn.Softmax()
self.mul = P.Mul()
self.add = P.Add()
self.sub = P.Sub()
self.div = P.Div()
self.assign = P.Assign()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 2, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
param_c = np.full((1,), 16, dtype=np.float32)
self.param_c = Parameter(Tensor(param_c), name='c')
def construct(self, x, y):
self.assign(self.param_a, x + self.param_a)
y = self.add(y, self.param_b)
if self.param_b == y - self.param_a:
self.param_c = self.div(self.param_c, self.param_b)
while self.param_a > x:
self.param_c = self.param_a + 2
x = x + 1
y = self.softmax(self.param_c)
self.param_b = self.sub(y, self.param_b)
x = self.mul(self.param_b, self.param_c)
for _ in range(0, 4):
x = self.sub(x, 3)
y = y + self.param_b
self.param_a = x + y
z = self.relu(y + self.param_a)
return z
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([11], mstype.int32)
y = Tensor([7], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_after_while_in_if_net = ForAfterWhileInIfNet()
net = GradNet(for_after_while_in_if_net)
graph_forward_res = for_after_while_in_if_net(x, y)
graph_backward_res = net(x, y)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_after_while_in_if_net = ForAfterWhileInIfNet()
net = GradNet(for_after_while_in_if_net)
pynative_forward_res = for_after_while_in_if_net(x, y)
pynative_backward_res = net(x, y)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

93
tests/st/control/inner/test_322_for_after_while_in_for.py Normal file
View File

@ -0,0 +1,93 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_for_after_while_in_for():
class ForAfterWhileInForNet(nn.Cell):
def __init__(self):
super().__init__()
self.relu = nn.ReLU()
self.softmax = nn.Softmax()
self.mul = P.Mul()
self.add = P.Add()
self.sub = P.Sub()
self.div = P.Div()
self.assign = P.Assign()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 2, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
param_c = np.full((1,), 16, dtype=np.float32)
self.param_c = Parameter(Tensor(param_c), name='c')
def construct(self, x, y):
self.assign(self.param_a, x + self.param_a)
y = self.add(y, self.param_b)
for _ in range(0, 3):
self.param_b = self.add(self.param_c, self.param_b)
while self.param_c > x:
self.param_b = self.param_a + 2
x = x + 1
y = self.softmax(self.param_c) + self.param_a
self.param_b = self.sub(y, self.param_b)
x = self.mul(self.param_b, self.param_c)
for _ in range(0, 4):
x = self.mul(x, 3)
y = y + self.param_b
x = self.relu(self.param_c)
self.param_a = x - y
z = y + self.param_b
return z
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([11], mstype.int32)
y = Tensor([7], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_after_while_in_for_net = ForAfterWhileInForNet()
net = GradNet(for_after_while_in_for_net)
graph_forward_res = for_after_while_in_for_net(x, y)
graph_backward_res = net(x, y)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_after_while_in_for_net = ForAfterWhileInForNet()
net = GradNet(for_after_while_in_for_net)
pynative_forward_res = for_after_while_in_for_net(x, y)
pynative_backward_res = net(x, y)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

88
tests/st/control/inner/test_331_for_after_for_in_while.py Normal file
View File

@ -0,0 +1,88 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_for_after_for_in_while():
class ForAfterForInWhileNet(nn.Cell):
def __init__(self):
super().__init__()
self.relu = nn.ReLU()
self.softmax = nn.Softmax()
self.mul = P.Mul()
self.add = P.Add()
self.sub = P.Sub()
self.div = P.Div()
self.assign = P.Assign()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 2, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
param_c = np.full((1,), 16, dtype=np.float32)
self.param_c = Parameter(Tensor(param_c), name='c')
def construct(self, x, y):
while self.param_c > x:
self.param_b = self.add(self.param_c, self.param_b)
for _ in range(0, 20):
self.param_b = self.param_a + 2
self.param_c = self.param_c - 1
x = x + 2
y = self.softmax(self.param_c) + self.param_a
self.param_b = self.sub(y, self.param_b)
x = self.mul(self.param_b, self.param_a)
for _ in range(0, 4):
x = self.mul(x, 3)
y = y + self.param_b
x = self.relu(self.param_c)
self.param_a = x - y
z = y + self.param_b
return z
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([11], mstype.int32)
y = Tensor([7], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_after_for_in_while_net = ForAfterForInWhileNet()
net = GradNet(for_after_for_in_while_net)
graph_forward_res = for_after_for_in_while_net(x, y)
graph_backward_res = net(x, y)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_after_for_in_while_net = ForAfterForInWhileNet()
net = GradNet(for_after_for_in_while_net)
pynative_forward_res = for_after_for_in_while_net(x, y)
pynative_backward_res = net(x, y)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res

84
tests/st/control/inner/test_332_for_after_for_in_for.py Normal file
View File

@ -0,0 +1,84 @@
# 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
from mindspore import context
from mindspore import Tensor, nn
from mindspore.common.parameter import Parameter
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
grad_all = C.GradOperation(get_all=True)
context.set_context(device_target="Ascend")
def test_for_after_for_in_for():
class ForAfterForInForNet(nn.Cell):
def __init__(self):
super().__init__()
self.relu = nn.ReLU()
self.softmax = nn.Softmax()
self.mul = P.Mul()
self.add = P.Add()
self.sub = P.Sub()
self.div = P.Div()
self.assign = P.Assign()
param_a = np.full((1,), 5, dtype=np.float32)
self.param_a = Parameter(Tensor(param_a), name='a')
param_b = np.full((1,), 2, dtype=np.float32)
self.param_b = Parameter(Tensor(param_b), name='b')
param_c = np.full((1,), 20, dtype=np.float32)
self.param_c = Parameter(Tensor(param_c), name='c')
def construct(self, x, y):
for _ in range(0, 4):
self.param_b = self.add(self.param_c, self.param_b)
for _ in range(0, 8):
self.param_b = self.param_a + j
self.param_c = self.param_a * self.param_b
for _ in range(0, 3):
y = y + self.param_b
x = self.relu(self.param_c * 3)
self.param_a = x - y
z = y + self.param_b
return z
class GradNet(nn.Cell):
def __init__(self, net):
super(GradNet, self).__init__()
self.net = net
def construct(self, *inputs):
return grad_all(self.net)(*inputs)
x = Tensor([11], mstype.int32)
y = Tensor([7], mstype.int32)
# graph mode
context.set_context(mode=context.GRAPH_MODE)
for_after_for_in_for_net = ForAfterForInForNet()
net = GradNet(for_after_for_in_for_net)
graph_forward_res = for_after_for_in_for_net(x, y)
graph_backward_res = net(x, y)
# pynative mode
context.set_context(mode=context.PYNATIVE_MODE)
for_after_for_in_for_net = ForAfterForInForNet()
net = GradNet(for_after_for_in_for_net)
pynative_forward_res = for_after_for_in_for_net(x, y)
pynative_backward_res = net(x, y)
assert graph_forward_res == pynative_forward_res
assert graph_backward_res == pynative_backward_res