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

pylint waring clean

This commit is contained in:
jinyaohui 2020-05-13 11:30:27 +08:00
parent a2d5ad5abe
commit 26fd75895d
108 changed files with 815 additions and 393 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
tests/perf_test/bert/test_bert_train.py
View File

@ -18,13 +18,13 @@
# pylint: disable=missing-docstring, arguments-differ, W0612
import os
import mindspore.common.dtype as mstype
import mindspore.context as context
from mindspore import Tensor
from mindspore.nn.optim import AdamWeightDecayDynamicLR
from mindspore.model_zoo.Bert_NEZHA import BertConfig, BertNetworkWithLoss, BertTrainOneStepCell, \
BertTrainOneStepWithLossScaleCell
from mindspore.nn.wrap.loss_scale import FixedLossScaleUpdateCell
from mindspore.nn.optim import AdamWeightDecayDynamicLR
from mindspore.train.loss_scale_manager import DynamicLossScaleManager
from ...dataset_mock import MindData
from ...ops_common import nn, np, batch_tuple_tensor, build_construct_graph

3
tests/perf_test/mindrecord/imagenet/imagenet_to_mindrecord.py
View File

@ -13,7 +13,6 @@
# limitations under the License.
# ============================================================================
"""use ImageNetToMR tool generate mindrecord"""
import os
from mindspore.mindrecord import ImageNetToMR
IMAGENET_MAP_FILE = "../../../ut/data/mindrecord/testImageNetDataWhole/labels_map.txt"
@ -21,6 +20,7 @@ IMAGENET_IMAGE_DIR = "../../../ut/data/mindrecord/testImageNetDataWhole/images"
MINDRECORD_FILE = "./imagenet.mindrecord"
PARTITION_NUMBER = 16
def imagenet_to_mindrecord():
imagenet_transformer = ImageNetToMR(IMAGENET_MAP_FILE,
IMAGENET_IMAGE_DIR,
@ -28,5 +28,6 @@ def imagenet_to_mindrecord():
PARTITION_NUMBER)
imagenet_transformer.transform()
if __name__ == '__main__':
imagenet_to_mindrecord()

7
tests/perf_test/mindrecord/imagenet/imagenet_to_tfrecord.py
View File

@ -15,6 +15,7 @@
"""generate tfrecord"""
import collections
import os
import tensorflow as tf
IMAGENET_MAP_FILE = "../../../ut/data/mindrecord/testImageNetDataWhole/labels_map.txt"
@ -22,6 +23,7 @@ IMAGENET_IMAGE_DIR = "../../../ut/data/mindrecord/testImageNetDataWhole/images"
TFRECORD_FILE = "./imagenet.tfrecord"
PARTITION_NUMBER = 16
def get_imagenet_filename_label_pic(map_file, image_dir):
"""
Get data from imagenet.
@ -69,18 +71,22 @@ def get_imagenet_filename_label_pic(map_file, image_dir):
continue
yield str(file_name), int(label), image_bytes
def create_int_feature(values):
feature = tf.train.Feature(int64_list=tf.train.Int64List(value=[values]))
return feature
def create_string_feature(values):
feature = tf.train.Feature(bytes_list=tf.train.BytesList(value=[bytes(values, encoding='utf-8')]))
return feature
def create_bytes_feature(values):
feature = tf.train.Feature(bytes_list=tf.train.BytesList(value=[values]))
return feature
def imagenet_to_tfrecord():
writers = []
for i in range(PARTITION_NUMBER):
@ -109,5 +115,6 @@ def imagenet_to_tfrecord():
print("Write {} total examples".format(total_written))
if __name__ == '__main__':
imagenet_to_tfrecord()

12
tests/perf_test/mindrecord/imagenet/perf_read_imagenet.py
View File

@ -14,17 +14,20 @@
# ============================================================================
"""test dataset performance about mindspore.MindDataset, mindspore.TFRecordDataset, tf.data.TFRecordDataset"""
import time
import mindspore.dataset as ds
from mindspore.mindrecord import FileReader
import tensorflow as tf
import mindspore.dataset as ds
from mindspore.mindrecord import FileReader
print_step = 5000
def print_log(count):
if count % print_step == 0:
print("Read {} rows ...".format(count))
def use_filereader(mindrecord):
start = time.time()
columns_list = ["data", "label"]
@ -38,6 +41,7 @@ def use_filereader(mindrecord):
end = time.time()
print("Read by FileReader - total rows: {}, cost time: {}s".format(num_iter, end - start))
def use_minddataset(mindrecord):
start = time.time()
columns_list = ["data", "label"]
@ -51,6 +55,7 @@ def use_minddataset(mindrecord):
end = time.time()
print("Read by MindDataset - total rows: {}, cost time: {}s".format(num_iter, end - start))
def use_tfrecorddataset(tfrecord):
start = time.time()
columns_list = ["data", "label"]
@ -66,8 +71,10 @@ def use_tfrecorddataset(tfrecord):
end = time.time()
print("Read by TFRecordDataset - total rows: {}, cost time: {}s".format(num_iter, end - start))
def use_tensorflow_tfrecorddataset(tfrecord):
start = time.time()
def _parse_record(example_photo):
features = {
'file_name': tf.io.FixedLenFeature([], tf.string),
@ -87,6 +94,7 @@ def use_tensorflow_tfrecorddataset(tfrecord):
end = time.time()
print("Read by TensorFlow TFRecordDataset - total rows: {}, cost time: {}s".format(num_iter, end - start))
if __name__ == '__main__':
# use MindDataset
mindrecord = './imagenet.mindrecord00'

10
tests/perf_test/test_lenet.py
View File

@ -18,15 +18,14 @@
import numpy as np
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor
from mindspore.model_zoo.lenet import LeNet
from mindspore import context
import mindspore.ops.composite as C
from mindspore import Tensor
from mindspore import context
from mindspore.common.api import _executor
from mindspore.model_zoo.lenet import LeNet
context.set_context(mode=context.GRAPH_MODE)
batch_size = 1
channel = 1
height = 32
@ -36,6 +35,7 @@ num_class = 10
class LeNetGrad(nn.Cell):
"""Backward of LeNet"""
def __init__(self, network):
super(LeNetGrad, self).__init__()
self.grad_op = C.grad_all_with_sens

3
tests/perf_test/test_resnet_infer.py
View File

@ -17,10 +17,11 @@
import numpy as np
from mindspore.common.api import _executor
from mindspore import Tensor
from mindspore.common.api import _executor
from .resnet_example import resnet50
def test_compile():
net = resnet50()
inp = Tensor(np.ones([1, 3, 224, 224]).astype(np.float32))

4
tests/perf_test/test_resnet_pynative.py
View File

@ -20,9 +20,9 @@
import numpy as np
from mindspore import Tensor
from ..train_step_wrap import train_step_without_opt
from .resnet_example import resnet50
from ..vm_impl import *
from ..train_step_wrap import train_step_without_opt
def test_resnet50_pynative():
net = train_step_without_opt(resnet50())

6
tests/perf_test/test_resnet_train.py
View File

@ -17,13 +17,15 @@
import numpy as np
from mindspore.common.api import _executor
import mindspore.context as context
from mindspore import Tensor
from ..train_step_wrap import train_step_with_loss_warp
from mindspore.common.api import _executor
from .resnet_example import resnet50
from ..train_step_wrap import train_step_with_loss_warp
context.set_context(mode=context.GRAPH_MODE)
def test_train_step():
net = train_step_with_loss_warp(resnet50())
net.set_train()

13
tests/train_step_wrap.py
View File

@ -16,15 +16,15 @@
train step wrap
"""
import mindspore.nn as nn
from mindspore.ops import functional as F
from mindspore import ParameterTuple
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore import Parameter, ParameterTuple
class TrainStepWrap(nn.Cell):
"""
TrainStepWrap definition
"""
def __init__(self, network):
super(TrainStepWrap, self).__init__()
self.network = network
@ -39,10 +39,12 @@ class TrainStepWrap(nn.Cell):
grads = self.grad(self.network, weights)(x, label)
return self.optimizer(grads)
class NetWithLossClass(nn.Cell):
"""
NetWithLossClass definition
"""
def __init__(self, network):
super(NetWithLossClass, self).__init__(auto_prefix=False)
self.loss = nn.SoftmaxCrossEntropyWithLogits()
@ -61,6 +63,7 @@ class TrainStepWrap2(nn.Cell):
"""
TrainStepWrap2 definition
"""
def __init__(self, network, sens):
super(TrainStepWrap2, self).__init__()
self.network = network
@ -76,13 +79,16 @@ class TrainStepWrap2(nn.Cell):
grads = self.grad(self.network, weights)(x, self.sens)
return self.optimizer(grads)
def train_step_with_sens(network, sens):
return TrainStepWrap2(network, sens)
class TrainStepWrapWithoutOpt(nn.Cell):
"""
TrainStepWrapWithoutOpt definition
"""
def __init__(self, network):
super(TrainStepWrapWithoutOpt, self).__init__()
self.network = network
@ -93,5 +99,6 @@ class TrainStepWrapWithoutOpt(nn.Cell):
grads = self.grad(self.network, self.weights)(x, label)
return grads
def train_step_without_opt(network):
return TrainStepWrapWithoutOpt(NetWithLossClass(network))

8
tests/ut/python/dtype/test_dictionary.py
View File

@ -28,6 +28,7 @@ context.set_context(mode=context.GRAPH_MODE)
def Xtest_arg_dict():
class DictNet(Cell):
"""DictNet definition"""
def __init__(self):
super(DictNet, self).__init__()
self.max = P.Maximum()
@ -48,6 +49,7 @@ def Xtest_arg_dict():
def test_const_dict():
class DictNet(Cell):
"""DictNet1 definition"""
def __init__(self):
super(DictNet, self).__init__()
self.max = P.Maximum()
@ -58,6 +60,7 @@ def test_const_dict():
a = self.max(self.dictionary["x"], self.dictionary["y"])
b = self.min(self.dictionary["x"], self.dictionary["y"])
return a + b
net = DictNet()
net()
@ -65,6 +68,7 @@ def test_const_dict():
def test_dict_set_or_get_item():
class DictNet(Cell):
"""DictNet1 definition"""
def __init__(self):
super(DictNet, self).__init__()
self.dict_ = {"x": 1, "y": 2}
@ -91,6 +95,7 @@ def test_dict_set_or_get_item():
def test_dict_set_or_get_item_2():
class DictNet(Cell):
"""DictNet1 definition"""
def __init__(self):
super(DictNet, self).__init__()
@ -117,6 +122,7 @@ def test_dict_set_or_get_item_2():
def test_dict_set_or_get_item_3():
class DictNet(Cell):
"""DictNet1 definition"""
def __init__(self):
super(DictNet, self).__init__()
self.dict_ = {"x": Tensor(np.ones([2, 2, 3], np.float32)), "y": 1}
@ -130,5 +136,3 @@ def test_dict_set_or_get_item_3():
net = DictNet()
assert net() == Tensor(np.ones([4, 2, 3], np.float32))

1
tests/ut/python/dtype/test_hypermap.py
View File

@ -13,7 +13,6 @@
# limitations under the License.
# ============================================================================
import numpy as np
import pytest
from mindspore import Tensor, context
from mindspore.nn import Cell

1
tests/ut/python/exec/__init__.py
View File

@ -15,6 +15,7 @@
"""setup for pytest"""
import mindspore.context as context
# pylint: disable=unused-argument
def setup_module(module):
context.set_context(mode=context.GRAPH_MODE)

1
tests/ut/python/exec/resnet_example.py
View File

@ -16,6 +16,7 @@
resnet50 example
"""
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.ops import operations as P

33
tests/ut/python/exec/test_AssignAdd.py
View File

@ -16,19 +16,21 @@
test assign add
"""
import numpy as np
import mindspore.nn as nn
from mindspore.ops import operations as P
from mindspore.common.initializer import initializer
from mindspore import Tensor, Parameter
import mindspore as ms
from ..ut_filter import non_graph_engine
from mindspore.common.api import _executor
import mindspore.context as context
import pytest
import mindspore.nn as nn
from mindspore import Tensor, Parameter
from mindspore.common.initializer import initializer
from mindspore.ops import operations as P
from ..ut_filter import non_graph_engine
context.set_context(mode=context.GRAPH_MODE)
class Net(nn.Cell):
"""Net definition"""
def __init__(self):
super(Net, self).__init__()
self.AssignAdd = P.AssignAdd()
@ -39,18 +41,19 @@ class Net(nn.Cell):
out = self.AssignAdd(self.inputdata, x)
return out
@non_graph_engine
def test_AssignAdd_1():
"""test AssignAdd 1"""
import mindspore.context as context
context.set_context(mode=context.GRAPH_MODE)
net = Net()
x = Tensor(np.ones([1]).astype(np.int64)*100)
x = Tensor(np.ones([1]).astype(np.int64) * 100)
print("MyPrintResult dataX:", x)
result = net(x)
print("MyPrintResult data::", result)
expect = np.ones([1]).astype(np.int64)*101
expect = np.ones([1]).astype(np.int64) * 101
diff = result.asnumpy() - expect
print("MyPrintExpect:", expect)
@ -58,18 +61,19 @@ def test_AssignAdd_1():
error = np.ones(shape=[1]) * 1.0e-3
assert np.all(diff < error)
@non_graph_engine
def test_AssignAdd_2():
"""test AssignAdd 2"""
import mindspore.context as context
context.set_context(mode=context.GRAPH_MODE)
net = Net()
x = Tensor(np.ones([1]).astype(np.int64)*102)
x = Tensor(np.ones([1]).astype(np.int64) * 102)
print("MyPrintResult dataX:", x)
result = net(x)
print("MyPrintResult data::", result.asnumpy())
expect = np.ones([1]).astype(np.int64)*103
expect = np.ones([1]).astype(np.int64) * 103
diff = result.asnumpy() - expect
print("MyPrintExpect:", expect)
@ -77,8 +81,10 @@ def test_AssignAdd_2():
error = np.ones(shape=[1]) * 1.0e-3
assert np.all(diff < error)
class AssignAddNet(nn.Cell):
"""Net definition"""
def __init__(self):
super(AssignAddNet, self).__init__()
self.AssignAdd = P.AssignAdd()
@ -89,9 +95,10 @@ class AssignAddNet(nn.Cell):
z1 = self.AssignAdd(self.inputdata, self.one)
return z1
@non_graph_engine
def test_assignadd_scalar_cast():
net = AssignAddNet()
x = Tensor(np.ones([1]).astype(np.int64)*102)
#_executor.compile(net, 1)
x = Tensor(np.ones([1]).astype(np.int64) * 102)
# _executor.compile(net, 1)
result = net(x)

1
tests/ut/python/exec/test_activation.py
View File

@ -14,6 +14,7 @@
# ============================================================================
""" test Activations """
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
from ..ut_filter import non_graph_engine

4
tests/ut/python/exec/test_assign_sub.py
View File

@ -16,15 +16,17 @@
test assign sub
"""
import numpy as np
import mindspore.context as context
import mindspore.nn as nn
import mindspore.ops.operations as P
from mindspore import Tensor
import mindspore.context as context
from mindspore.common.initializer import initializer
from mindspore.common.parameter import Parameter
context.set_context(mode=context.GRAPH_MODE)
class Net(nn.Cell):
def __init__(self):
super(Net, self).__init__()

1
tests/ut/python/exec/test_batchnorm.py
View File

@ -14,6 +14,7 @@
# ============================================================================
"""ut for batchnorm layer"""
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
from ..ut_filter import non_graph_engine

7
tests/ut/python/exec/test_bias_add.py
View File

@ -14,14 +14,17 @@
# ============================================================================
""" test BiasAdd """
import numpy as np
import mindspore.nn as nn
from mindspore.ops import operations as P
from mindspore.common.initializer import initializer
from mindspore import Tensor, Parameter
from mindspore.common.initializer import initializer
from mindspore.ops import operations as P
from ..ut_filter import non_graph_engine
class Net(nn.Cell):
"""Net definition"""
def __init__(self,
output_channels,
bias_init='zeros',

3
tests/ut/python/exec/test_conv.py
View File

@ -14,6 +14,7 @@
# ============================================================================
"""test conv"""
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
from ..ut_filter import non_graph_engine
@ -25,6 +26,7 @@ out_channels = 64
class Net(nn.Cell):
"""Net definition"""
def __init__(self,
cin,
cout,
@ -70,6 +72,7 @@ def test_compile2():
output = net(input_data)
print(output.asnumpy())
@non_graph_engine
def test_compile3():
net = Net(3, 1, (3, 3), weight_init='ONES')

3
tests/ut/python/exec/test_dense.py
View File

@ -14,12 +14,15 @@
# ============================================================================
""" test Dense """
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
from ..ut_filter import non_graph_engine
class Net(nn.Cell):
"""Net definition"""
def __init__(self,
input_channels,
output_channels,

3
tests/ut/python/exec/test_eval.py
View File

@ -14,11 +14,12 @@
# ============================================================================
"""test eval"""
import numpy as np
import mindspore as ms
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor
from mindspore import context
from mindspore.common.api import _executor
from ..ut_filter import non_graph_engine

2
tests/ut/python/exec/test_flatten.py
View File

@ -16,8 +16,8 @@
import numpy as np
import mindspore.nn as nn
from mindspore.ops import operations as P
from mindspore import Tensor
from mindspore.ops import operations as P
from ..ut_filter import non_graph_engine

4
tests/ut/python/exec/test_pooling.py
View File

@ -15,12 +15,12 @@
"""
test pooling api
"""
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
class MaxNet(nn.Cell):
"""MaxNet definition"""
def __init__(self,
kernel_size,
stride=None):

2
tests/ut/python/exec/test_softmax.py
View File

@ -16,9 +16,11 @@
test softmax api
"""
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
class Net(nn.Cell):
def __init__(self, dim):
super(Net, self).__init__()

2
tests/ut/python/exec/test_tensor_add.py
View File

@ -14,10 +14,12 @@
# ============================================================================
""" test TensorAdd """
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.ops import operations as P
class Net(nn.Cell):
def __init__(self):
super(Net, self).__init__()

17
tests/ut/python/exec/test_train.py
View File

@ -14,29 +14,33 @@
# ============================================================================
""" test model train """
import numpy as np
import mindspore.nn as nn
from mindspore.ops import operations as P
from mindspore.common.initializer import initializer
from mindspore import Tensor, Parameter, Model
from mindspore.common.initializer import initializer
from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
from mindspore.nn.optim import Momentum
from mindspore.ops import operations as P
# fn is a funcation use i as input
def lr_gen(fn, epoch_size):
for i in range(epoch_size):
yield fn(i)
def me_train_tensor(net, input_np, label_np, epoch_size=2):
"""me_train_tensor"""
loss = SoftmaxCrossEntropyWithLogits(is_grad=False, sparse=True, reduction="mean")
opt = Momentum(filter(lambda x: x.requires_grad, net.get_parameters()), lr_gen(lambda i: 0.1, epoch_size), 0.9, 0.01, 1024)
opt = Momentum(filter(lambda x: x.requires_grad, net.get_parameters()), lr_gen(lambda i: 0.1, epoch_size), 0.9,
0.01, 1024)
Model(net, loss, opt)
_network = nn.WithLossCell(net, loss)
_train_net = nn.TrainOneStepCell(_network, opt)
_train_net.set_train()
label_np = np.argmax(label_np, axis=-1).astype(np.int32)
for epoch in range(0, epoch_size):
print(f"epoch %d"%(epoch))
print(f"epoch %d" % (epoch))
_train_net(Tensor(input_np), Tensor(label_np))
@ -52,6 +56,7 @@ def test_bias_add(test_with_simu):
class Net(nn.Cell):
"""Net definition"""
def __init__(self,
output_channels,
bias_init='zeros',
@ -87,6 +92,7 @@ def test_conv(test_with_simu):
class Net(nn.Cell):
"Net definition"""
def __init__(self,
cin,
cout,
@ -116,6 +122,7 @@ def test_net():
class Net(nn.Cell):
"""Net definition"""
def __init__(self):
super(Net, self).__init__()
Tensor(np.ones([64, 3, 7, 7]).astype(np.float32) * 0.01)
@ -141,6 +148,7 @@ def test_net():
label_np = np.ones([32, 12]).astype(np.int32)
me_train_tensor(net, input_np, label_np)
def test_bn():
"""test_bn"""
import mindspore.context as context
@ -151,6 +159,7 @@ def test_bn():
class Net(nn.Cell):
"""Net definition"""
def __init__(self, cin, cout):
super(Net, self).__init__()
self.bn = nn.BatchNorm2d(cin)

6
tests/ut/python/exec/test_train_with_lars.py
View File

@ -23,6 +23,7 @@ from mindspore.ops import composite as C
from mindspore.ops import functional as F
from mindspore.ops import operations as P
def get_reordered_parameters(parameters):
"""get_reordered_parameters"""
# put the bias parameter to the end
@ -36,12 +37,15 @@ def get_reordered_parameters(parameters):
reordered_params = tuple(non_bias_param + bias_param)
return len(non_bias_param), len(reordered_params), reordered_params
def get_net_trainable_reordered_params(net):
params = net.trainable_params()
return get_reordered_parameters(params)
class TrainOneStepWithLarsCell(nn.Cell):
"""TrainOneStepWithLarsCell definition"""
def __init__(self, network, optimizer, sens=1.0):
super(TrainOneStepWithLarsCell, self).__init__(auto_prefix=False)
self.network = network
@ -66,11 +70,13 @@ class TrainOneStepWithLarsCell(nn.Cell):
new_grads = lars_grads + bias_grads
return F.depend(loss, self.optimizer(new_grads))
# fn is a funcation use i as input
def lr_gen(fn, epoch_size):
for i in range(epoch_size):
yield fn(i)
def me_train_tensor(net, input_np, label_np, epoch_size=2):
"""me_train_tensor"""
loss = SoftmaxCrossEntropyWithLogits(is_grad=False, sparse=True)

3
tests/ut/python/ir/test_dtype.py
View File

@ -14,12 +14,14 @@
# ============================================================================
"""test_dtype"""
from dataclasses import dataclass
import numpy as np
import pytest
import mindspore as ms
from mindspore.common import dtype
def test_dtype_to_nptype():
"""test_dtype2nptype"""
assert ms.dtype_to_nptype(ms.bool_) == np.bool_
@ -59,6 +61,7 @@ def test_dtype_to_pytype():
@dataclass
class Foo:
x: int
def inf(self):
return self.x

47
tests/ut/python/ir/test_tensor.py
View File

@ -25,25 +25,27 @@ import mindspore as ms
import mindspore.common.api as me
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.parameter import Parameter
from mindspore.common.initializer import initializer
from mindspore.common.parameter import Parameter
from ..ut_filter import non_graph_engine
ndarr = np.ones((2, 3))
def test_tensor_flatten():
with pytest.raises(AttributeError):
lst = [1, 2, 3, 4,]
lst = [1, 2, 3, 4, ]
tensor_list = ms.Tensor(lst, ms.float32)
tensor_list = tensor_list.Flatten()
print(tensor_list)
def test_tensor_list():
lst = [[1.0, 2.0, 1.0], [1.0, 10.0, 9.0]]
tensor_list = ms.Tensor(lst, ms.float32)
print(tensor_list)
def test_tensor():
"""test_tensor"""
t1 = ms.Tensor(ndarr)
@ -63,6 +65,7 @@ def test_tensor():
assert isinstance(t4, ms.Tensor)
assert t4.dtype() == ms.int64
def test_tensor_type_float16():
t_float16 = ms.Tensor(np.array([[1, 2, 3], [4, 5, 6]], dtype=np.float16))
assert isinstance(t_float16, ms.Tensor)
@ -107,6 +110,7 @@ def test_tensor_type_float64_user_define():
assert t_float64.shape() == (2, 3)
assert t_float64.dtype() == ms.float64
def test_tensor_type_bool():
# init a tensor with bool type
ts_bool_array = ms.Tensor(np.zeros([2, 3], np.bool), ms.bool_)
@ -122,6 +126,7 @@ def test_tensor_type_bool():
assert t_bool_array.shape() == (2, 3)
assert t_bool_array.dtype() == ms.bool_
def test_tensor_type_int8():
t_int8_array = ms.Tensor(np.array([[1, 2, 3], [4, 5, 6]], dtype=np.int8))
assert isinstance(t_int8_array, ms.Tensor)
@ -154,6 +159,7 @@ def test_tensor_type_int64():
assert t_int64.shape() == (2, 3)
assert t_int64.dtype() == ms.int64
def test_tensor_type_uint8():
t_uint8_array = ms.Tensor(np.array([[1, 2, 3], [4, 5, 6]], dtype=np.uint8))
assert isinstance(t_uint8_array, ms.Tensor)
@ -181,6 +187,7 @@ def test_tensor_type_uint64():
assert t_uint64.shape() == (2, 3)
assert t_uint64.dtype() == ms.uint64
def test_set_type():
t = ms.Tensor(ndarr)
t.set_dtype(ms.float32)
@ -202,15 +209,17 @@ def test_sub():
z = x - y
assert isinstance(z, ms.Tensor)
@non_graph_engine
def test_div():
x = ms.Tensor(np.array([[2,6,10],[12, 4, 8]]).astype(np.float32))
y = ms.Tensor(np.array([[2,2,5],[6, 1, 2]]).astype(np.float32))
x = ms.Tensor(np.array([[2, 6, 10], [12, 4, 8]]).astype(np.float32))
y = ms.Tensor(np.array([[2, 2, 5], [6, 1, 2]]).astype(np.float32))
z = x / y
z2 = x / 2
assert isinstance(z, ms.Tensor)
assert isinstance(z2, ms.Tensor)
@non_graph_engine
def test_parameter():
x = Parameter(initializer(1, [1], ms.float32), name="beta1_power")
@ -220,6 +229,7 @@ def test_parameter():
class Net(nn.Cell):
"""Net definition"""
def __init__(self, dim):
super(Net, self).__init__()
self.dim = dim
@ -266,6 +276,7 @@ def test_tensor_contiguous():
assert True, rt_f.flags['C_CONTIGUOUS']
print("rt_f flags = ", rt_f.flags)
def test_tensor_contiguous2():
input_data = np.random.randn(32, 112, 112, 3).astype(np.float32)
input_me = input_data.transpose(0, 3, 1, 2)
@ -274,36 +285,43 @@ def test_tensor_contiguous2():
out_f = tensor_f_float32.asnumpy()
print("out_f flags = ", out_f.flags)
def test_tensor_input_string():
with pytest.raises(TypeError):
input_data = 'ccc'
ms.Tensor(input_data)
def test_tensor_input_tuple_string():
with pytest.raises(TypeError):
input_data = (2, 3, '4', 5)
ms.Tensor(input_data)
def test_tensor_input_list_string():
with pytest.raises(TypeError):
input_data = [[2, 3, '4', 5], [1, 2, 3, 4]]
ms.Tensor(input_data)
def test_tensor_input_none():
with pytest.raises(TypeError):
input_data = None
ms.Tensor(input_data, np.int64)
# pylint: disable=no-value-for-parameter
def test_tensor_input_empty():
with pytest.raises(TypeError):
ms.Tensor()
def test_tensor_input_ndarray_str():
with pytest.raises(TypeError):
inp = np.array(["88", 2, 4])
ms.Tensor(inp)
def test_tensor_input_ndarray_bool():
inp = np.array([True, 2, 4])
ms.Tensor(inp)
@ -311,86 +329,103 @@ def test_tensor_input_ndarray_bool():
inp = np.array([False, 2, 4])
ms.Tensor(inp)
def test_tensor_input_ndarray_complex():
with pytest.raises(TypeError):
inp = np.array([20j, 2, 4])
ms.Tensor(inp)
def test_tensor_input_ndarray_none():
with pytest.raises(TypeError):
inp = np.array([None, 2, 4])
ms.Tensor(inp)
def test_tensor_input_ndarray_dict():
with pytest.raises(TypeError):
inp = {'a': 6, 'b': 7}
ms.Tensor(inp)
def test_tensor_input_np_nan():
with pytest.raises(TypeError):
input_data = (1, 2, 3, np.nan)
ms.Tensor(input_data, np.int64)
def test_tensor_input_tuple_inf():
with pytest.raises(TypeError):
input_data = (1, 2, 3, float("inf"))
ms.Tensor(input_data, np.int64)
def test_tensor_input_dict():
with pytest.raises(TypeError):
input_data = {'a': 6, 'b': 7}
ms.Tensor(input_data, np.int64)
def test_tensor_input_complex():
with pytest.raises(TypeError):
input_data = (1, 2j, 3)
ms.Tensor(input_data, np.int64)
def test_tensor_dtype_np_float():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.float)
ms.Tensor(input_data, np.float)
def test_tensor_dtype_np_float16():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.float16)
ms.Tensor(input_data, np.float16)
def test_tensor_dtype_np_float32():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.float32)
ms.Tensor(input_data, np.float32)
def test_tensor_dtype_np_float64():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.float64)
ms.Tensor(input_data, np.float64)
def test_tensor_dtype_np_int():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.int)
ms.Tensor(input_data, np.int)
def test_tensor_dtype_np_int8():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.int8)
ms.Tensor(input_data, np.int8)
def test_tensor_dtype_np_int16():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.int16)
ms.Tensor(input_data, np.int16)
def test_tensor_dtype_np_int32():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.int32)
ms.Tensor(input_data, np.int32)
def test_tensor_dtype_np_int64():
with pytest.raises(TypeError):
input_data = np.random.randn(32, 112, 112, 3).astype(np.int64)
ms.Tensor(input_data, np.int64)
def test_tensor_dtype_fp32_to_bool():
with pytest.raises(RuntimeError):
input = np.random.randn(2, 3, 4, 5).astype(np.float32)
@ -399,7 +434,7 @@ def test_tensor_dtype_fp32_to_bool():
def test_tensor_operation():
x = Tensor(np.ones((3,3)) * 4)
x = Tensor(np.ones((3, 3)) * 4)
res = x + 1
assert np.all(res.asnumpy() == np.ones((3, 3)) * 5)
res = 1 + x

8
tests/ut/python/ir/test_tensor_py.py
View File

@ -14,10 +14,11 @@
# ============================================================================
"""test tensor py"""
import numpy as np
import mindspore as ms
from mindspore.common.api import _executor
from mindspore.nn import Cell
from mindspore.ops import operations as P
from mindspore.common.api import _executor
import mindspore as ms
from ..ut_filter import non_graph_engine
@ -93,6 +94,7 @@ def test_float():
def test_tensor_method_sub():
"""test_tensor_method_sub"""
class Net(Cell):
def __init__(self):
super(Net, self).__init__()
@ -111,6 +113,7 @@ def test_tensor_method_sub():
def test_tensor_method_mul():
"""test_tensor_method_mul"""
class Net(Cell):
def __init__(self):
super(Net, self).__init__()
@ -129,6 +132,7 @@ def test_tensor_method_mul():
def test_tensor_method_div():
"""test_tensor_method_div"""
class Net(Cell):
def __init__(self):
super(Net, self).__init__()

33
tests/ut/python/keep_order/test_keep_order.py
View File

@ -13,24 +13,25 @@
# limitations under the License.
# ============================================================================
import numpy as np
from mindspore.common.api import ms_function
from mindspore.common.tensor import Tensor
from mindspore.ops import operations as P
import mindspore.ops.functional as F
import mindspore.context as context
from mindspore.ops import composite as C
from mindspore.ops.composite import core
from mindspore.common import dtype as mstype
import mindspore.nn as nn
import mindspore.ops.functional as F
from mindspore.common import dtype as mstype
from mindspore.common.tensor import Tensor
from mindspore.ops import composite as C
from mindspore.ops import operations as P
context.set_context(mode=context.GRAPH_MODE)
add1 = P.TensorAdd()
mul1 = P.MatMul()
add2 = P.TensorAdd()
def add(x, y):
return add1(x, y)
class Func(nn.Cell):
def __init__(self):
super(Func, self).__init__()
@ -48,7 +49,10 @@ class Func(nn.Cell):
out = F.depend(out, clear)
return out
grad_s = C.GradOperation('grad_with_sens', get_all=True, sens_param=True)
class Net(nn.Cell):
def __init__(self):
super(Net, self).__init__()
@ -69,6 +73,7 @@ class Net(nn.Cell):
out = F.depend(out, clear)
return out
def test_add():
x = Tensor(np.ones([3, 3]).astype(np.float32))
y = Tensor(np.ones([3, 3]).astype(np.float32))
@ -76,6 +81,7 @@ def test_add():
func.add_flags(has_effect=True)
func(x, y)
def test_sens():
x = Tensor(np.ones([3, 3]).astype(np.float32))
y = Tensor(np.ones([3, 3]).astype(np.float32))
@ -84,6 +90,7 @@ def test_sens():
net.add_flags(has_effect=True)
out = net(x, y, sens)
class Net_hyper(nn.Cell):
def __init__(self):
super(Net_hyper, self).__init__()
@ -105,6 +112,7 @@ class Net_hyper(nn.Cell):
out = F.depend(out, clear)
return out
def test_hyper_add():
x = Tensor(np.ones([3, 3]).astype(np.float32))
y = Tensor(np.ones([3, 3]).astype(np.float32))
@ -113,10 +121,11 @@ def test_hyper_add():
net.add_flags(has_effect=True)
out = net(x, y, sens)
def test_keep_order_io_effect_exception_return_dtype():
class Net(nn.Cell):
def __init__(self):
super().__init__()
super().__init__()
self.alloc_status = P.NPUAllocFloatStatus()
self.get_status = P.NPUGetFloatStatus()
self.clear_status = P.NPUClearFloatStatus()
@ -126,16 +135,16 @@ def test_keep_order_io_effect_exception_return_dtype():
self.neg = P.Neg()
self.add_flags(has_effect=True)
def construct(self, x):
def construct(self, x):
init = self.alloc_status()
self.clear_status(init)
res = self.sub(x, self.neg(x))
res = self.sub(x, self.neg(x))
self.get_status(init)
dtype = self.dtype(res)
return dtype
value = 655
value = 655
data = np.full((8, 5, 3, 1), value, dtype=np.float16)
x = Tensor(data, dtype=mstype.float16)
net = Net()
data = net(x)
data = net(x)

17
tests/ut/python/metrics/test_accuracy.py
View File

@ -14,10 +14,13 @@
# ============================================================================
"""test accuracy"""
import math
import numpy as np
import pytest
from mindspore.nn.metrics import Accuracy
from mindspore import Tensor
from mindspore.nn.metrics import Accuracy
def test_classification_accuracy():
"""test_classification_accuracy"""
@ -29,8 +32,9 @@ def test_classification_accuracy():
metric.update(x, y)
accuracy = metric.eval()
accuracy2 = metric(x, y2)
assert math.isclose(accuracy, 2/3)
assert math.isclose(accuracy2, 2/3)
assert math.isclose(accuracy, 2 / 3)
assert math.isclose(accuracy2, 2 / 3)
def test_multilabel_accuracy():
x = Tensor(np.array([[0, 1, 0, 1], [1, 0, 1, 1], [0, 0, 0, 1]]))
@ -39,7 +43,8 @@ def test_multilabel_accuracy():
metric.clear()
metric.update(x, y)
accuracy = metric.eval()
assert accuracy == 1/3
assert accuracy == 1 / 3
def test_shape_accuracy():
x = Tensor(np.array([[0, 1, 0, 1], [1, 0, 1, 1], [0, 0, 0, 1]]))
@ -49,6 +54,7 @@ def test_shape_accuracy():
with pytest.raises(ValueError):
metric.update(x, y)
def test_shape_accuracy2():
x = Tensor(np.array([[0, 1, 0, 1], [1, 0, 1, 1], [0, 0, 0, 1]]))
y = Tensor(np.array([0, 1, 1, 1]))
@ -57,6 +63,7 @@ def test_shape_accuracy2():
with pytest.raises(ValueError):
metric.update(x, y)
def test_shape_accuracy3():
x = Tensor(np.array([[0.2, 0.5], [0.3, 0.1], [0.9, 0.6]]))
y = Tensor(np.array([[1, 0, 1], [1, 1, 1]]))
@ -65,6 +72,7 @@ def test_shape_accuracy3():
with pytest.raises(ValueError):
metric.update(x, y)
def test_shape_accuracy4():
x = Tensor(np.array([[0.2, 0.5], [0.3, 0.1], [0.9, 0.6]]))
y = Tensor(np.array(1))
@ -73,6 +81,7 @@ def test_shape_accuracy4():
with pytest.raises(ValueError):
metric.update(x, y)
def test_type_accuracy():
with pytest.raises(TypeError):
Accuracy('test')

8
tests/ut/python/metrics/test_error.py
View File

@ -14,10 +14,12 @@
# ============================================================================
"""test error"""
import math
import numpy as np
import pytest
from mindspore.nn.metrics import MAE, MSE
from mindspore import Tensor
from mindspore.nn.metrics import MAE, MSE
def test_MAE():
@ -27,7 +29,7 @@ def test_MAE():
error.clear()
error.update(x, y)
result = error.eval()
assert math.isclose(result, 0.15/4)
assert math.isclose(result, 0.15 / 4)
def test_input_MAE():
@ -52,7 +54,7 @@ def test_MSE():
error.clear()
error.update(x, y)
result = error.eval()
assert math.isclose(result, 0.0125/4)
assert math.isclose(result, 0.0125 / 4)
def test_input_MSE():

11
tests/ut/python/metrics/test_fbeta.py
View File

@ -13,11 +13,11 @@
# limitations under the License.
# ============================================================================
# """test_fbeta"""
import math
import numpy as np
import pytest
from mindspore.nn.metrics import get_metric_fn, Fbeta
from mindspore import Tensor
from mindspore.nn.metrics import get_metric_fn, Fbeta
def test_classification_fbeta():
@ -32,9 +32,9 @@ def test_classification_fbeta():
fbeta_mean = metric.eval(True)
fbeta2 = metric(x, y2)
assert np.allclose(fbeta, np.array([2/3, 2/3]))
assert np.allclose(fbeta2, np.array([2/3, 2/3]))
assert np.allclose(fbeta_mean, 2/3)
assert np.allclose(fbeta, np.array([2 / 3, 2 / 3]))
assert np.allclose(fbeta2, np.array([2 / 3, 2 / 3]))
assert np.allclose(fbeta_mean, 2 / 3)
def test_fbeta_update1():
@ -46,6 +46,7 @@ def test_fbeta_update1():
with pytest.raises(ValueError):
metric.update(x, y)
def test_fbeta_update2():
x1 = Tensor(np.array([[0.2, 0.5, 0.7], [0.3, 0.1, 0.2], [0.9, 0.6, 0.5]]))
y1 = Tensor(np.array([1, 0, 2]))

3
tests/ut/python/metrics/test_loss.py
View File

@ -15,8 +15,9 @@
"""test loss"""
import numpy as np
import pytest
from mindspore.nn.metrics import Loss
from mindspore import Tensor
from mindspore.nn.metrics import Loss
def test_loss_inputs_error():

8
tests/ut/python/metrics/test_metric_factory.py
View File

@ -14,9 +14,11 @@
# ============================================================================
"""test_metric_factory"""
import math
import numpy as np
from mindspore.nn.metrics import get_metric_fn
from mindspore import Tensor
from mindspore.nn.metrics import get_metric_fn
def test_classification_accuracy():
@ -26,7 +28,7 @@ def test_classification_accuracy():
metric.clear()
metric.update(x, y)
accuracy = metric.eval()
assert math.isclose(accuracy, 2/3)
assert math.isclose(accuracy, 2 / 3)
def test_classification_accuracy_by_alias():
@ -36,7 +38,7 @@ def test_classification_accuracy_by_alias():
metric.clear()
metric.update(x, y)
accuracy = metric.eval()
assert math.isclose(accuracy, 2/3)
assert math.isclose(accuracy, 2 / 3)
def test_classification_precision():

8
tests/ut/python/metrics/test_precision.py
View File

@ -14,10 +14,12 @@
# ============================================================================
"""test_precision"""
import math
import numpy as np
import pytest
from mindspore.nn.metrics import Precision
from mindspore import Tensor
from mindspore.nn.metrics import Precision
def test_classification_precision():
@ -43,7 +45,7 @@ def test_multilabel_precision():
metric.update(x, y)
precision = metric.eval()
assert np.equal(precision, np.array([1, 2/3, 1])).all()
assert np.equal(precision, np.array([1, 2 / 3, 1])).all()
def test_average_precision():
@ -54,7 +56,7 @@ def test_average_precision():
metric.update(x, y)
precision = metric.eval(True)
assert math.isclose(precision, (1 + 2/3 + 1) / 3)
assert math.isclose(precision, (1 + 2 / 3 + 1) / 3)
def test_num_precision():

8
tests/ut/python/metrics/test_recall.py
View File

@ -14,10 +14,12 @@
# ============================================================================
"""test recall"""
import math
import numpy as np
import pytest
from mindspore.nn.metrics import Recall
from mindspore import Tensor
from mindspore.nn.metrics import Recall
def test_classification_recall():
@ -43,7 +45,7 @@ def test_multilabel_recall():
metric.update(x, y)
recall = metric.eval()
assert np.equal(recall, np.array([2/3, 2/3, 1])).all()
assert np.equal(recall, np.array([2 / 3, 2 / 3, 1])).all()
def test_average_recall():
@ -54,7 +56,7 @@ def test_average_recall():
metric.update(x, y)
recall = metric.eval(True)
assert math.isclose(recall, (2/3 + 2/3 + 1) / 3)
assert math.isclose(recall, (2 / 3 + 2 / 3 + 1) / 3)
def test_num_recall():

16
tests/ut/python/metrics/test_topk.py
View File

@ -14,10 +14,12 @@
# ============================================================================
"""test topk"""
import math
import numpy as np
import pytest
from mindspore.nn.metrics import TopKCategoricalAccuracy, Top1CategoricalAccuracy, Top5CategoricalAccuracy
from mindspore import Tensor
from mindspore.nn.metrics import TopKCategoricalAccuracy, Top1CategoricalAccuracy, Top5CategoricalAccuracy
def test_type_topk():
@ -54,8 +56,8 @@ def test_topk():
topk.update(x, y)
result = topk.eval()
result2 = topk(x, y2)
assert math.isclose(result, 2/3)
assert math.isclose(result2, 2/3)
assert math.isclose(result, 2 / 3)
assert math.isclose(result2, 2 / 3)
def test_zero_topk():
@ -79,8 +81,8 @@ def test_top1():
topk.update(x, y)
result = topk.eval()
result2 = topk(x, y2)
assert math.isclose(result, 1/3)
assert math.isclose(result2, 1/3)
assert math.isclose(result, 1 / 3)
assert math.isclose(result2, 1 / 3)
def test_top5():
@ -97,5 +99,5 @@ def test_top5():
topk.update(x, y)
result = topk.eval()
result2 = topk(x, y2)
assert math.isclose(result, 2/3)
assert math.isclose(result2, 2/3)
assert math.isclose(result, 2 / 3)
assert math.isclose(result2, 2 / 3)

1
tests/ut/python/model/__init__.py
View File

@ -15,6 +15,7 @@
"""setup for pytest"""
import mindspore.context as context
# pylint: disable=unused-argument
def setup_module(module):
context.set_context(mode=context.GRAPH_MODE)

4
tests/ut/python/model/res18_example.py
View File

@ -17,10 +17,10 @@ resnet50 example
"""
import numpy as np
from mindspore.common.api import _executor
import mindspore.nn as nn # pylint: disable=C0414
from mindspore import Tensor
from mindspore.common.api import _executor
from mindspore.ops.operations import TensorAdd
import mindspore.nn as nn # pylint: disable=C0414
from ...train_step_wrap import train_step_with_loss_warp

5
tests/ut/python/model/test_bert.py
View File

@ -15,9 +15,8 @@
""" test bert cell """
import numpy as np
import pytest
from mindspore import Model
from mindspore.nn.optim import AdamWeightDecay
from mindspore.model_zoo.Bert_NEZHA import BertConfig, BertModel, BertNetworkWithLoss, BertTrainOneStepCell
from mindspore.model_zoo.Bert_NEZHA import BertConfig, BertModel
from ....dataset_mock import MindData

53
tests/ut/python/model/test_bert_cell.py
View File

@ -14,26 +14,30 @@
# ============================================================================
""" test bert of graph compile """
import functools
import numpy as np
import mindspore.common.dtype as mstype
import mindspore.nn as nn
import mindspore.ops.composite as C
from mindspore.common.initializer import TruncatedNormal
from mindspore.common.parameter import ParameterTuple
from mindspore.common.tensor import Tensor
from mindspore.model_zoo.Bert_NEZHA import BertPretrainingLoss, GetNextSentenceOutput
from mindspore.model_zoo.Bert_NEZHA.bert_for_pre_training import ClipGradients
from mindspore.model_zoo.Bert_NEZHA.bert_model import BertConfig, \
EmbeddingLookup, EmbeddingPostprocessor, BertOutput, RelaPosMatrixGenerator, \
RelaPosEmbeddingsGenerator, SaturateCast, BertAttention, BertSelfAttention, \
BertEncoderCell, BertTransformer, CreateAttentionMaskFromInputMask, BertModel
from mindspore.nn.layer.basic import Norm
from mindspore.model_zoo.Bert_NEZHA import BertPretrainingLoss, GetNextSentenceOutput
import mindspore.nn as nn
from mindspore.common.initializer import TruncatedNormal
from mindspore.common.parameter import ParameterTuple
from mindspore.nn.optim import AdamWeightDecay, AdamWeightDecayDynamicLR
from mindspore.model_zoo.Bert_NEZHA.bert_for_pre_training import ClipGradients
import mindspore.ops.composite as C
from mindspore.ops import functional as F
from ....ops_common import convert
from ....mindspore_test_framework.mindspore_test import mindspore_test
from ....mindspore_test_framework.pipeline.forward.compile_forward import pipeline_for_compile_forward_ge_graph_for_case_by_case_config
from ....mindspore_test_framework.pipeline.gradient.compile_gradient import pipeline_for_compile_grad_ge_graph_for_case_by_case_config
from ....mindspore_test_framework.pipeline.forward.compile_forward import \
pipeline_for_compile_forward_ge_graph_for_case_by_case_config
from ....mindspore_test_framework.pipeline.gradient.compile_gradient import \
pipeline_for_compile_grad_ge_graph_for_case_by_case_config
from ....ops_common import convert
def bert_trans():
"""bert_trans"""
@ -53,10 +57,12 @@ def bert_trans():
net.set_train()
return net
def set_train(net):
net.set_train()
return net
class NetForAdam(nn.Cell):
def __init__(self):
super(NetForAdam, self).__init__()
@ -66,8 +72,10 @@ class NetForAdam(nn.Cell):
x = self.dense(x)
return x
class TrainStepWrapForAdam(nn.Cell):
"""TrainStepWrapForAdam definition"""
def __init__(self, network):
super(TrainStepWrapForAdam, self).__init__()
self.network = network
@ -81,8 +89,10 @@ class TrainStepWrapForAdam(nn.Cell):
grads = self.clip_gradients(grads, 1, 1.0)
return self.optimizer(grads)
class TrainStepWrapForAdamDynamicLr(nn.Cell):
"""TrainStepWrapForAdamDynamicLr definition"""
def __init__(self, network):
super(TrainStepWrapForAdamDynamicLr, self).__init__()
self.network = network
@ -95,16 +105,19 @@ class TrainStepWrapForAdamDynamicLr(nn.Cell):
grads = C.grad_by_list_with_sens(self.network, weights)(x, self.sens)
return self.optimizer(grads)
class TempC2Wrap(nn.Cell):
def __init__(self, op, c1=None, c2=None,):
def __init__(self, op, c1=None, c2=None, ):
super(TempC2Wrap, self).__init__()
self.op = op
self.c1 = c1
self.c2 = c2
def construct(self, x1):
x = self.op(x1, self.c1, self.c2)
return x
test_case_cell_ops = [
('Norm_keepdims', {
'block': Norm(keep_dims=True),
@ -373,7 +386,7 @@ test_case_cell_ops = [
'block': set_train(nn.Dense(in_channels=768,
out_channels=3072,
activation='gelu',
weight_init=TruncatedNormal(0.02),)),
weight_init=TruncatedNormal(0.02), )),
'desc_inputs': [[3, 768]],
'desc_bprop': [[3, 3072]]}),
('GetNextSentenceOutput', {
@ -396,26 +409,28 @@ test_case_cell_ops = [
'block': TempC2Wrap(ClipGradients(), 1, 1.0),
'desc_inputs': [tuple(convert(shp) for shp in [[1], [1], [1]])],
'skip': ['backward', 'exec']}),
]
]
test_case = functools.reduce(lambda x, y: x+y, [test_case_cell_ops])
test_case = functools.reduce(lambda x, y: x + y, [test_case_cell_ops])
# use -k to select certain testcast
# pytest tests/python/ops/test_ops.py::test_backward -k LayerNorm
test_exec_case = filter(lambda x: 'skip' not in x[1] or
'exec' not in x[1]['skip'], test_case)
'exec' not in x[1]['skip'], test_case)
test_backward_exec_case = filter(lambda x: 'skip' not in x[1] or
'backward' not in x[1]['skip'] and 'backward_exec'
not in x[1]['skip'], test_case)
'backward' not in x[1]['skip'] and 'backward_exec'
not in x[1]['skip'], test_case)
test_check_gradient_case = filter(lambda x: 'skip' not in x[1] or
'backward' not in x[1]['skip'] and 'backward_exec'
not in x[1]['skip'], test_case)
'backward' not in x[1]['skip'] and 'backward_exec'
not in x[1]['skip'], test_case)
@mindspore_test(pipeline_for_compile_forward_ge_graph_for_case_by_case_config)
def test_exec():
return test_exec_case
@mindspore_test(pipeline_for_compile_grad_ge_graph_for_case_by_case_config)
def test_backward_exec():
return test_backward_exec_case

11
tests/ut/python/model/test_lenet.py
View File

@ -15,16 +15,19 @@
"""test lenet"""
import numpy as np
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor
from mindspore.ops import operations as P
import mindspore.context as context
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.api import _executor
from mindspore.ops import operations as P
from ....train_step_wrap import train_step_with_loss_warp, train_step_with_sens
context.set_context(mode=context.GRAPH_MODE)
class LeNet5(nn.Cell):
"""LeNet5 definition"""
def __init__(self):
super(LeNet5, self).__init__()
self.conv1 = nn.Conv2d(1, 6, 5, pad_mode='valid')

3
tests/ut/python/model/test_lenet_core_after_exception.py
View File

@ -14,8 +14,8 @@
# ============================================================================
"""test_lenet_core_after_exception"""
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore.common.tensor import Tensor
@ -25,6 +25,7 @@ from ....train_step_wrap import train_step_with_loss_warp
class LeNet5(nn.Cell):
"""LeNet5 definition"""
def __init__(self):
super(LeNet5, self).__init__()
self.conv1 = nn.Conv2d(3, 6, 5, pad_mode="valid")

22
tests/ut/python/model/test_mix_precision.py
View File

@ -15,23 +15,24 @@
"""test_mix_precision"""
import numpy as np
import mindspore.nn as nn
import mindspore.common.dtype as mstype
import mindspore.nn as nn
from mindspore import Tensor, context
from mindspore.common import ParameterTuple
from mindspore.common.api import _executor
from mindspore.common.parameter import Parameter
from mindspore.common import ParameterTuple
from mindspore import Tensor, context
from mindspore.ops import operations as P
from mindspore.ops import composite as C
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.nn import Momentum
from ....train_step_wrap import train_step_with_loss_warp
from tests.ops_common import convert
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from mindspore.train.parallel_utils import ParallelMode
from tests.ops_common import convert
from ....train_step_wrap import train_step_with_loss_warp
class LeNet5(nn.Cell):
"""LeNet5"""
def __init__(self):
super(LeNet5, self).__init__()
self.conv1 = nn.Conv2d(1, 6, 5, pad_mode='valid')
@ -148,10 +149,13 @@ def test_cast():
"""test grad of PReLU, which cause AddN(generated by grad) fail"""
class IRBlockZ(nn.Cell):
def __init__(self, inplanes, planes):
super(IRBlockZ, self).__init__()
self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=3, stride=1, pad_mode="same", group=1, has_bias=False, dilation=1)
self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=3, stride=1, pad_mode="same", group=1, has_bias=False,
dilation=1)
self.act_layer = nn.PReLU(planes)
def construct(self, x):

1
tests/ut/python/model/test_utils.py
View File

@ -14,6 +14,7 @@
# ============================================================================
"""test_dataset_utils"""
import pytest
import mindspore as ms
from mindspore.train._utils import _construct_tensor_list

1
tests/ut/python/nn/__init__.py
View File

@ -15,6 +15,7 @@
"""setup for pytest"""
import mindspore.context as context
# pylint: disable=unused-argument
def setup_module(module):
context.set_context(mode=context.GRAPH_MODE)

6
tests/ut/python/nn/optim/test_adam.py
View File

@ -15,16 +15,18 @@
""" test adam """
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor, Parameter
from mindspore.common.api import _executor
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.ops import operations as P
from mindspore.nn.optim import AdamWeightDecay, AdamWeightDecayDynamicLR
from mindspore.ops import operations as P
class Net(nn.Cell):
""" Net definition """
def __init__(self):
super(Net, self).__init__()
self.weight = Parameter(Tensor(np.ones([64, 10]).astype(np.float32)), name="weight")

2
tests/ut/python/nn/optim/test_ftrl.py
View File

@ -15,6 +15,7 @@
""" test FTRL """
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor, Parameter
from mindspore.common.api import _executor
@ -47,4 +48,3 @@ def test_ftrl():
net_with_loss = WithLossCell(net, loss)
train_network = TrainOneStepCell(net_with_loss, optimizer)
_executor.compile(train_network, inputs, label)

7
tests/ut/python/nn/optim/test_lamb.py
View File

@ -14,16 +14,18 @@
# ============================================================================
""" test lamb """
import numpy as np
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor, Parameter
from mindspore.common.api import _executor
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.ops import operations as P
from mindspore.nn.optim import Lamb
from mindspore.ops import operations as P
class Net(nn.Cell):
""" Net definition """
def __init__(self):
super(Net, self).__init__()
self.weight = Parameter(Tensor(np.ones([64, 10]).astype(np.float32)), name="weight")
@ -38,6 +40,7 @@ class Net(nn.Cell):
class NetWithoutWeight(nn.Cell):
""" NetWithoutWeight definition """
def __init__(self):
super(NetWithoutWeight, self).__init__()
self.matmul = P.MatMul()

10
tests/ut/python/nn/optim/test_lars.py
View File

@ -12,15 +12,17 @@
# See the License for the specific language governing permissions and
# limitations under the License.
# ============================================================================
from collections import Counter
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor, Parameter
from mindspore.common import dtype as mstype
from mindspore.common.api import _executor
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.nn.optim import LARS, Momentum
from mindspore.ops import operations as P
from mindspore.common import dtype as mstype
from collections import Counter
def multisteplr(total_steps, milestone, base_lr=0.9, gamma=0.1, dtype=mstype.float32):
@ -56,7 +58,7 @@ def test_lars_multi_step_lr():
lr = multisteplr(10, [2, 6])
SGD = Momentum(net.trainable_params(), lr, 0.9)
optimizer = LARS(SGD, epsilon=1e-08, hyperpara=0.02, decay_filter=lambda x: 'bn' not in x.name,
lars_filter=lambda x: 'bn' not in x.name)
lars_filter=lambda x: 'bn' not in x.name)
net_with_loss = WithLossCell(net, loss)
train_network = TrainOneStepCell(net_with_loss, optimizer)
@ -77,4 +79,4 @@ def test_lars_float_lr():
net_with_loss = WithLossCell(net, loss)
train_network = TrainOneStepCell(net_with_loss, optimizer)
_executor.compile(train_network, inputs, label)
_executor.compile(train_network, inputs, label)

14
tests/ut/python/nn/optim/test_lr_schedule.py
View File

@ -14,19 +14,17 @@
# ============================================================================
""" test_lr_schedule """
import numpy as np
from mindspore.nn import Cell
from mindspore.ops.operations import BiasAdd, MatMul
from mindspore import Parameter, ParameterTuple, Tensor
from mindspore.nn import WithLossCell
from mindspore.nn.optim import Momentum
from mindspore.nn import SoftmaxCrossEntropyWithLogits
from mindspore.ops.composite import grad_by_list
from mindspore.ops import functional as F
from mindspore.nn import Cell
from mindspore.nn.optim import Optimizer
from mindspore.ops.composite import grad_by_list
from mindspore.ops.operations import BiasAdd, MatMul
class Net(Cell):
""" Net definition """
def __init__(self):
super(Net, self).__init__()
self.weight = Parameter(Tensor(np.ones([64, 10])), name="weight")
@ -41,6 +39,7 @@ class Net(Cell):
class _TrainOneStepCell(Cell):
""" _TrainOneStepCell definition """
def __init__(self, network, optimizer):
"""
Append an optimizer to the training network after that the construct
@ -67,4 +66,3 @@ class _TrainOneStepCell(Cell):
if self.lr_schedule:
self.schedule.update_lr(*args)
return self.optimizer(grads)

6
tests/ut/python/nn/optim/test_momentum.py
View File

@ -14,16 +14,18 @@
# ============================================================================
""" test momentum """
import numpy as np
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor, Parameter
from mindspore.common.api import _executor
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.ops import operations as P
from mindspore.nn.optim import Momentum
from mindspore.ops import operations as P
class Net(nn.Cell):
""" Net definition """
def __init__(self):
super(Net, self).__init__()
self.weight = Parameter(Tensor(np.ones([64, 10]).astype(np.float32)), name="weight")

9
tests/ut/python/nn/optim/test_optimizer.py
View File

@ -15,9 +15,10 @@
""" test optimizer """
import numpy as np
import pytest
from mindspore import Tensor
from mindspore.nn.optim import Optimizer, SGD, Adam, AdamWeightDecay, AdamWeightDecayDynamicLR
from mindspore.common.parameter import Parameter
from mindspore.nn.optim import Optimizer, SGD, Adam, AdamWeightDecay, AdamWeightDecayDynamicLR
class IterableObjc:
@ -30,6 +31,7 @@ class IterableObjc:
params = IterableObjc()
class TestOptimizer():
def test_init(self):
Optimizer(0.5, params)
@ -44,6 +46,7 @@ class TestOptimizer():
class TestAdam():
""" TestAdam definition """
def test_init(self):
Adam(params, learning_rate=1e-3, beta1=0.9, beta2=0.999, eps=1e-8, use_locking=False,
use_nesterov=False, weight_decay=0.0, loss_scale=1.0)
@ -58,6 +61,7 @@ class TestAdam():
class TestSGD():
""" TestSGD definition """
def test_init(self):
with pytest.raises(ValueError):
SGD(params, learning_rate=0.1, momentum=-0.1, dampening=0, weight_decay=0, nesterov=False)
@ -68,6 +72,7 @@ class TestSGD():
class TestNullParam():
""" TestNullParam definition """
def test_optim_init(self):
with pytest.raises(ValueError):
Optimizer(0.1, None)
@ -84,8 +89,10 @@ class TestNullParam():
with pytest.raises(ValueError):
SGD(None)
class TestUnsupportParam():
""" TestUnsupportParam definition """
def test_optim_init(self):
with pytest.raises(ValueError):
Optimizer(0.1, (1, 2, 3))

7
tests/ut/python/nn/optim/test_rmsprop.py
View File

@ -15,16 +15,18 @@
""" test adam """
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor, Parameter
from mindspore.common.api import _executor
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.ops import operations as P
from mindspore.nn.optim import RMSProp
from mindspore.ops import operations as P
class Net(nn.Cell):
""" Net definition """
def __init__(self):
super(Net, self).__init__()
self.weight = Parameter(Tensor(np.ones([64, 10]).astype(np.float32)), name="weight")
@ -59,4 +61,3 @@ def test_rmsprop_e():
with pytest.raises(TypeError):
RMSProp(net.get_parameters(), momentum=1, learning_rate=0.1)

4
tests/ut/python/nn/test_activation.py
View File

@ -14,11 +14,13 @@
# ============================================================================
""" test Activations """
import numpy as np
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor
from mindspore.common.api import _executor
from ..ut_filter import non_graph_engine
class SoftmaxNet(nn.Cell):
def __init__(self, dim):
super(SoftmaxNet, self).__init__()

5
tests/ut/python/nn/test_batchnorm.py
View File

@ -17,8 +17,8 @@ import numpy as np
import pytest
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor, Parameter
from mindspore.common.api import _executor
def test_bn_pars_valid1():
@ -62,11 +62,12 @@ class GroupNet(nn.Cell):
def __init__(self):
super(GroupNet, self).__init__()
self.group_bn = nn.GroupNorm()
def construct(self, x):
return self.group_bn(x)
def test_compile_groupnorm():
net = nn.GroupNorm(16, 64)
input_data = Tensor(np.random.rand(1,64,256,256).astype(np.float32))
input_data = Tensor(np.random.rand(1, 64, 256, 256).astype(np.float32))
_executor.compile(net, input_data)

5
tests/ut/python/nn/test_cell.py
View File

@ -15,11 +15,10 @@
""" test cell """
import numpy as np
import pytest
import mindspore.context as context
import mindspore.nn as nn
from mindspore import Tensor, Parameter
from mindspore.common.api import _executor
from ..ut_filter import non_graph_engine
class ModA(nn.Cell):
@ -90,7 +89,7 @@ class ConvNet(nn.Cell):
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode="same")
self.flatten = nn.Flatten()
self.fc = nn.Dense(
int(ConvNet.image_h*ConvNet.image_w*ConvNet.output_ch/(4*4)),
int(ConvNet.image_h * ConvNet.image_w * ConvNet.output_ch / (4 * 4)),
num_classes)
def construct(self, x):

5
tests/ut/python/nn/test_cell_wrapper.py
View File

@ -12,14 +12,15 @@
# See the License for the specific language governing permissions and
# limitations under the License.
# ============================================================================
import pytest
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore import Tensor, Parameter
from mindspore.common import dtype as mstype
from mindspore.common.api import _executor
from mindspore.nn import TrainOneStepCell, WithLossCell, ParameterUpdate
from mindspore.nn.optim import Momentum
from mindspore.common import dtype as mstype
from mindspore.ops import operations as P

2
tests/ut/python/nn/test_checkparameter.py
View File

@ -14,10 +14,10 @@
# ============================================================================
""" test checkparameter """
import pytest
from mindspore._checkparam import check_int, check_int_positive, \
check_input_format, check_bool, twice
kernel_size = 5
kernel_size1 = twice(kernel_size)
assert kernel_size1 == (5, 5)

3
tests/ut/python/nn/test_clip_by_norm.py
View File

@ -14,11 +14,12 @@
# ============================================================================
""" test clip_by_norm """
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore import Tensor
from ..ut_filter import non_graph_engine
@non_graph_engine
def test_clip_by_norm():
clip_by_norm = nn.ClipByNorm()

6
tests/ut/python/nn/test_container.py
View File

@ -14,12 +14,13 @@
# ============================================================================
""" test container """
from collections import OrderedDict
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore import Tensor
weight = Tensor(np.ones([2, 2]))
conv2 = nn.Conv2d(3, 64, (3, 3), stride=2, padding=0)
@ -31,6 +32,7 @@ avg_pool = nn.AvgPool2d(kernel_size, stride)
class TestSequentialCell():
""" TestSequentialCell """
def test_SequentialCell_init(self):
m = nn.SequentialCell()
assert type(m).__name__ == 'SequentialCell'
@ -86,6 +88,7 @@ class TestSequentialCell():
class TestCellList():
""" TestCellList """
def test_init1(self):
cell_list = nn.CellList([conv2, avg_pool])
assert len(cell_list) == 2
@ -118,7 +121,6 @@ class TestCellList():
cell = item
assert type(cell).__name__ == 'AvgPool2d'
def test_add(self):
cell_list = nn.CellList([conv2, avg_pool])
cell_list += [conv2]

6
tests/ut/python/nn/test_dense.py
View File

@ -15,10 +15,11 @@
""" test nn.Dense """
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore.common.api import _executor
import mindspore.context as context
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.api import _executor
from ..ut_filter import non_graph_engine
@ -68,6 +69,7 @@ def test_dense_channels_error():
class Net(nn.Cell):
""" Net definition """
def __init__(self,
input_channels,
output_channels,

2
tests/ut/python/nn/test_dropout.py
View File

@ -15,12 +15,14 @@
""" Test Dropout """
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore import Tensor
from mindspore import context
context.set_context(device_target="Ascend")
def test_check_dropout_3():
Tensor(np.ones([20, 16, 50]).astype(np.int32))
with pytest.raises(ValueError):

14
tests/ut/python/nn/test_dynamic_lr.py
View File

@ -14,7 +14,7 @@
# ============================================================================
""" Test Dynamic Learning Rate """
import pytest
import mindspore
from mindspore.nn import dynamic_lr as dr
milestone = [10, 20, 30]
@ -29,8 +29,9 @@ min_lr = 0.01
max_lr = 0.1
power = 0.5
class TestInputs:
def test_milestone1(self):
def test_milestone1(self):
milestone1 = 1
with pytest.raises(TypeError):
dr.piecewise_constant_lr(milestone1, learning_rates)
@ -58,7 +59,7 @@ class TestInputs:
lr = True
with pytest.raises(TypeError):
dr.exponential_decay_lr(lr, decay_rate, total_step, step_per_epoch, decay_epoch)
with pytest.raises(TypeError):
dr.polynomial_decay_lr(lr, end_learning_rate, total_step, step_per_epoch, decay_epoch, power)
@ -71,7 +72,7 @@ class TestInputs:
dr.polynomial_decay_lr(lr, end_learning_rate, total_step, step_per_epoch, decay_epoch, power)
def test_end_learning_rate_type(self):
lr = True
lr = True
with pytest.raises(TypeError):
dr.polynomial_decay_lr(learning_rate, lr, total_step, step_per_epoch, decay_epoch, power)
@ -127,7 +128,7 @@ class TestInputs:
step_per_epoch1 = -1
with pytest.raises(ValueError):
dr.exponential_decay_lr(learning_rate, decay_rate, total_step, step_per_epoch1, decay_epoch)
with pytest.raises(ValueError):
dr.cosine_decay_lr(min_lr, max_lr, total_step, step_per_epoch1, decay_epoch)
@ -226,9 +227,10 @@ def test_cosine_decay():
lr = dr.cosine_decay_lr(min_lr, max_lr, total_step, step_per_epoch, decay_epoch)
assert len(lr) == total_step
def test_polynomial_decay():
lr1 = dr.polynomial_decay_lr(learning_rate, end_learning_rate, total_step, step_per_epoch, decay_epoch, power)
assert len(lr1) == total_step
lr2 = dr.polynomial_decay_lr(learning_rate, end_learning_rate, total_step, step_per_epoch, decay_epoch, power,
True)
True)
assert len(lr2) == total_step

9
tests/ut/python/nn/test_embedding.py
View File

@ -14,11 +14,10 @@
# ============================================================================
""" test_embedding """
import numpy as np
import pytest
from mindspore.model_zoo.Bert_NEZHA import EmbeddingLookup, EmbeddingPostprocessor
from mindspore import Tensor
from mindspore import dtype as mstype
from mindspore.model_zoo.Bert_NEZHA import EmbeddingLookup, EmbeddingPostprocessor
from ..ut_filter import non_graph_engine
@ -39,6 +38,7 @@ def test_check_embedding_lookup_2():
use_one_hot_embeddings=True)
m(Tensor(np.ones([128]), mstype.int32))
@non_graph_engine
def test_check_embedding_lookup_3():
m = EmbeddingLookup(vocab_size=32000,
@ -48,11 +48,12 @@ def test_check_embedding_lookup_3():
initializer_range=0.01)
m(Tensor(np.ones([128]), mstype.int32))
@non_graph_engine
def test_embedding_post_1():
m = EmbeddingPostprocessor(embedding_size=768,
embedding_shape=[1, 128, 768],
use_token_type=True)
embedding_shape=[1, 128, 768],
use_token_type=True)
m(Tensor(np.ones([128]), mstype.int32), Tensor(np.ones([1, 128, 768]), mstype.float32))

3
tests/ut/python/nn/test_flatten.py
View File

@ -16,9 +16,10 @@
test flatten api
"""
import numpy as np
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor
from mindspore.common.api import _executor
class Net(nn.Cell):

21
tests/ut/python/nn/test_image_gradients.py
View File

@ -15,14 +15,17 @@
""" test image gradients """
import numpy as np
import pytest
import mindspore.nn as nn
import mindspore.context as context
import mindspore.common.dtype as mstype
import mindspore.context as context
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.api import _executor
from mindspore.common.api import ms_function
context.set_context(device_target="Ascend")
class Net(nn.Cell):
def __init__(self):
super(Net, self).__init__()
@ -32,9 +35,10 @@ class Net(nn.Cell):
def construct(self, x):
return self.image_gradients(x)
def test_compile():
# input shape 1 x 1 x 2 x 2
image = Tensor(np.array([[[[1,2],[3,4]]]]), dtype=mstype.int32)
image = Tensor(np.array([[[[1, 2], [3, 4]]]]), dtype=mstype.int32)
net = Net()
_executor.compile(net, image)
@ -42,16 +46,17 @@ def test_compile():
def test_compile_multi_channel():
# input shape 4 x 2 x 2 x 2
dtype = mstype.int32
image = Tensor(np.array([[[[1,2],[3,4]], [[5,6],[7,8]]],
[[[3,5],[7,9]], [[11,13],[15,17]]],
[[[5,10],[15,20]], [[25,30],[35,40]]],
[[[10,20],[30,40]], [[50,60],[70,80]]]]), dtype=dtype)
image = Tensor(np.array([[[[1, 2], [3, 4]], [[5, 6], [7, 8]]],
[[[3, 5], [7, 9]], [[11, 13], [15, 17]]],
[[[5, 10], [15, 20]], [[25, 30], [35, 40]]],
[[[10, 20], [30, 40]], [[50, 60], [70, 80]]]]), dtype=dtype)
net = Net()
_executor.compile(net, image)
def test_invalid_5d_input():
dtype = mstype.float32
image = Tensor(np.random.random([4, 1, 16, 16, 1]), dtype=dtype)
net = Net()
with pytest.raises(ValueError):
_executor.compile(net, image)
_executor.compile(net, image)

6
tests/ut/python/nn/test_loss.py
View File

@ -14,12 +14,12 @@
# ============================================================================
""" test loss """
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.api import _executor
from ..ut_filter import non_graph_engine
import mindspore
def test_L1Loss():
loss = nn.L1Loss()
@ -60,5 +60,5 @@ def test_SoftmaxCrossEntropyExpand():
loss = nn.SoftmaxCrossEntropyExpand()
logits = Tensor(np.random.randint(0, 9, [100, 10]).astype(np.float32))
labels = Tensor(np.random.randint(0, 9, [10,]).astype(np.float32))
labels = Tensor(np.random.randint(0, 9, [10, ]).astype(np.float32))
_executor.compile(loss, logits, labels)

4
tests/ut/python/nn/test_lstm.py
View File

@ -14,6 +14,7 @@
# ============================================================================
""" test lstm """
import pytest
import mindspore.context as context
from mindspore import nn
from ..ut_filter import run_on_gpu
@ -22,6 +23,7 @@ from ....ops_common import convert
class LstmTestNet(nn.Cell):
""" LstmTestNet definition """
def __init__(self, input_size, hidden_size, num_layers, has_bias, batch_first, bidirectional):
super(LstmTestNet, self).__init__()
self.lstm = nn.LSTM(input_size=input_size,
@ -32,7 +34,6 @@ class LstmTestNet(nn.Cell):
bidirectional=bidirectional,
dropout=0.0)
def construct(self, inp, h0, c0):
return self.lstm(inp, (h0, c0))
@ -86,6 +87,7 @@ def test_compile(args):
out = net(*inputs)
print(f"out: {out}")
@run_on_gpu
@pytest.mark.parametrize('args', test_case_cell_ops, ids=lambda x: x[0])
def test_execute(args):

4
tests/ut/python/nn/test_nn_embedding.py
View File

@ -14,10 +14,11 @@
# ============================================================================
""" test nn embedding """
import numpy as np
from mindspore import Tensor
from mindspore.common import dtype
from mindspore.nn import Embedding
from mindspore.common.api import _executor
from mindspore.nn import Embedding
from ..ut_filter import non_graph_engine
@ -41,6 +42,7 @@ def test_check_embedding_3():
input_data = Tensor(np.ones([8, 128]), dtype.int32)
_executor.compile(net, input_data)
@non_graph_engine
def test_print_embedding():
net = Embedding(20000, 768, False)

11
tests/ut/python/nn/test_nn_pad.py
View File

@ -13,13 +13,12 @@
# limitations under the License.
# ============================================================================
""" test nn pad """
from mindspore import Tensor
from mindspore.ops import operations as P
import mindspore.nn as nn
from mindspore.ops.composite import GradOperation
from mindspore.common.api import ms_function
import numpy as np
import mindspore.context as context
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.api import ms_function
from mindspore.ops.composite import GradOperation
class Net(nn.Cell):

1
tests/ut/python/nn/test_norm.py
View File

@ -14,6 +14,7 @@
# ============================================================================
""" test norm """
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.api import _executor

21
tests/ut/python/nn/test_parameter.py
View File

@ -15,10 +15,11 @@
""" test parameter """
import numpy as np
import pytest
from mindspore import Tensor, Parameter, ParameterTuple
from mindspore._checkparam import _check_str_by_regular
from mindspore.common import dtype as mstype
from mindspore.common.initializer import initializer
from mindspore._checkparam import _check_str_by_regular
def test_parameter_init():
@ -30,7 +31,7 @@ def test_parameter_init():
def test_parameter_tuple_illegal():
p1 = Parameter(initializer(0, [1], mstype.int32), name="global_step1")
p2 = Parameter(initializer(0, [1], mstype.int32), name="global_step2")
plist = [p1,p2]
plist = [p1, p2]
plist2 = [p1, "str"]
ptuple = (p1, p2)
ptuple_str = ("2", "1")
@ -100,21 +101,21 @@ def test_parameter_init_illegal():
with pytest.raises(TypeError):
Parameter(tensor, name=data_str, requires_grad=data_tuple)
Parameter(tensor, name=data_str, requires_grad=data_bool,layerwise_parallel=data_bool)
Parameter(tensor, name=data_str, requires_grad=data_bool, layerwise_parallel=data_bool)
with pytest.raises(TypeError):
Parameter(tensor, name=data_str, requires_grad=data_bool,layerwise_parallel=dat)
Parameter(tensor, name=data_str, requires_grad=data_bool, layerwise_parallel=dat)
with pytest.raises(TypeError):
Parameter(tensor, name=data_str, requires_grad=data_bool,layerwise_parallel=tensor)
Parameter(tensor, name=data_str, requires_grad=data_bool, layerwise_parallel=tensor)
with pytest.raises(TypeError):
Parameter(tensor, name=data_str, requires_grad=data_bool,layerwise_parallel=data_none)
Parameter(tensor, name=data_str, requires_grad=data_bool, layerwise_parallel=data_none)
with pytest.raises(TypeError):
Parameter(tensor, name=data_str, requires_grad=data_bool,layerwise_parallel=data_str)
Parameter(tensor, name=data_str, requires_grad=data_bool, layerwise_parallel=data_str)
with pytest.raises(TypeError):
Parameter(tensor, name=data_str, requires_grad=data_bool,layerwise_parallel=data_int)
Parameter(tensor, name=data_str, requires_grad=data_bool, layerwise_parallel=data_int)
with pytest.raises(TypeError):
Parameter(tensor, name=data_str, requires_grad=data_bool,layerwise_parallel=data_list)
Parameter(tensor, name=data_str, requires_grad=data_bool, layerwise_parallel=data_list)
with pytest.raises(TypeError):
Parameter(tensor, name=data_str, requires_grad=data_bool,layerwise_parallel=data_tuple)
Parameter(tensor, name=data_str, requires_grad=data_bool, layerwise_parallel=data_tuple)
def test_check_str_by_regular():

7
tests/ut/python/nn/test_pooling.py
View File

@ -16,9 +16,10 @@
test pooling api
"""
import numpy as np
import mindspore.nn as nn
from mindspore.common.api import _executor
from mindspore import Tensor
from mindspore.common.api import _executor
class AvgNet(nn.Cell):
@ -40,6 +41,7 @@ def test_compile_avg():
class MaxNet(nn.Cell):
""" MaxNet definition """
def __init__(self,
kernel_size,
stride=None,
@ -68,7 +70,8 @@ class Avg1dNet(nn.Cell):
def construct(self, x):
return self.avg1d(x)
def test_avg1d():
net = Avg1dNet(6, 1)
input = Tensor(np.random.randint(0, 255, [1, 3, 6]).astype(np.float32))
_executor.compile(net, input)
_executor.compile(net, input)

11
tests/ut/python/nn/test_psnr.py
View File

@ -17,11 +17,11 @@ test psnr
"""
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common import dtype as mstype
from mindspore.common.api import _executor
from mindspore import Tensor
class PSNRNet(nn.Cell):
@ -40,6 +40,7 @@ def test_compile_psnr():
img2 = Tensor(np.random.random((8, 3, 16, 16)))
_executor.compile(net, img1, img2)
def test_compile_psnr_grayscale():
max_val = 255
net = PSNRNet(max_val)
@ -47,21 +48,25 @@ def test_compile_psnr_grayscale():
img2 = Tensor(np.random.randint(0, 256, (8, 1, 16, 16), np.uint8))
_executor.compile(net, img1, img2)
def test_psnr_max_val_negative():
max_val = -1
with pytest.raises(ValueError):
net = PSNRNet(max_val)
def test_psnr_max_val_bool():
max_val = True
with pytest.raises(TypeError):
net = PSNRNet(max_val)
def test_psnr_max_val_zero():
max_val = 0
with pytest.raises(ValueError):
net = PSNRNet(max_val)
def test_psnr_different_shape():
shape_1 = (8, 3, 16, 16)
shape_2 = (8, 3, 8, 8)
@ -71,6 +76,7 @@ def test_psnr_different_shape():
with pytest.raises(ValueError):
_executor.compile(net, img1, img2)
def test_psnr_different_dtype():
dtype_1 = mstype.float32
dtype_2 = mstype.float16
@ -80,6 +86,7 @@ def test_psnr_different_dtype():
with pytest.raises(TypeError):
_executor.compile(net, img1, img2)
def test_psnr_invalid_5d_input():
shape_1 = (8, 3, 16, 16)
shape_2 = (8, 3, 8, 8)

19
tests/ut/python/nn/test_ssim.py
View File

@ -17,10 +17,11 @@ test ssim
"""
import numpy as np
import pytest
import mindspore.nn as nn
import mindspore.common.dtype as mstype
from mindspore.common.api import _executor
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.api import _executor
class SSIMNet(nn.Cell):
@ -38,44 +39,53 @@ def test_compile():
img2 = Tensor(np.random.random((8, 3, 16, 16)))
_executor.compile(net, img1, img2)
def test_compile_grayscale():
max_val = 255
net = SSIMNet(max_val = max_val)
net = SSIMNet(max_val=max_val)
img1 = Tensor(np.random.randint(0, 256, (8, 1, 16, 16), np.uint8))
img2 = Tensor(np.random.randint(0, 256, (8, 1, 16, 16), np.uint8))
_executor.compile(net, img1, img2)
def test_ssim_max_val_negative():
max_val = -1
with pytest.raises(ValueError):
net = SSIMNet(max_val)
def test_ssim_max_val_bool():
max_val = True
with pytest.raises(TypeError):
net = SSIMNet(max_val)
def test_ssim_max_val_zero():
max_val = 0
with pytest.raises(ValueError):
net = SSIMNet(max_val)
def test_ssim_filter_size_float():
with pytest.raises(TypeError):
net = SSIMNet(filter_size=1.1)
def test_ssim_filter_size_zero():
with pytest.raises(ValueError):
net = SSIMNet(filter_size=0)
def test_ssim_filter_sigma_zero():
with pytest.raises(ValueError):
net = SSIMNet(filter_sigma=0.0)
def test_ssim_filter_sigma_negative():
with pytest.raises(ValueError):
net = SSIMNet(filter_sigma=-0.1)
def test_ssim_k1_k2_wrong_value():
with pytest.raises(ValueError):
net = SSIMNet(k1=1.1)
@ -95,6 +105,7 @@ def test_ssim_k1_k2_wrong_value():
with pytest.raises(ValueError):
net = SSIMNet(k2=-1.0)
def test_ssim_different_shape():
shape_1 = (8, 3, 16, 16)
shape_2 = (8, 3, 8, 8)
@ -104,6 +115,7 @@ def test_ssim_different_shape():
with pytest.raises(ValueError):
_executor.compile(net, img1, img2)
def test_ssim_different_dtype():
dtype_1 = mstype.float32
dtype_2 = mstype.float16
@ -113,6 +125,7 @@ def test_ssim_different_dtype():
with pytest.raises(TypeError):
_executor.compile(net, img1, img2)
def test_ssim_invalid_5d_input():
shape_1 = (8, 3, 16, 16)
shape_2 = (8, 3, 8, 8)

2
tests/ut/python/nn/test_structure_output.py
View File

@ -17,9 +17,9 @@ test_structure_output
"""
import numpy as np
import mindspore.ops.operations as P
from mindspore import Tensor, context
from mindspore.nn import Cell
import mindspore.ops.operations as P
from mindspore.ops.functional import depend
context.set_context(mode=context.GRAPH_MODE)

1
tests/ut/python/optimizer/__init__.py
View File

@ -13,4 +13,3 @@
# limitations under the License.
# ============================================================================
""" init vm impl """
from ....vm_impl import vm

48
tests/ut/python/optimizer/test_debug_location.py
View File

@ -13,27 +13,26 @@
# limitations under the License.
# ============================================================================
import numpy as np
import mindspore.nn as nn
import pytest
from mindspore import context
import mindspore.nn as nn
from mindspore import Tensor, Parameter
from mindspore.nn.wrap.cell_wrapper import WithLossCell
from mindspore.train.loss_scale_manager import FixedLossScaleManager, DynamicLossScaleManager
from mindspore.nn.wrap.loss_scale import TrainOneStepWithLossScaleCell
from mindspore.ops import operations as P
from mindspore.nn.optim import Momentum
from mindspore.ops import functional as F
from mindspore import context
from mindspore.common import dtype as mstype
from mindspore.train import Model
from ....dataset_mock import MindData
from mindspore.nn.optim import Lamb
from mindspore.ops._utils import _get_broadcast_shape
from mindspore.ops.primitive import Primitive, PrimitiveWithInfer, prim_attr_register
from mindspore.nn.optim import Momentum
from mindspore.nn.wrap.cell_wrapper import WithLossCell
from mindspore.nn.wrap.loss_scale import TrainOneStepWithLossScaleCell
from mindspore.ops import functional as F
from mindspore.ops import operations as P
from mindspore.ops._grad.grad_base import bprop_getters
from mindspore.ops._grad.grad_math_ops import binop_grad_common
from mindspore.ops._utils import _get_broadcast_shape
from mindspore.ops.primitive import PrimitiveWithInfer, prim_attr_register
from mindspore.train.loss_scale_manager import DynamicLossScaleManager
context.set_context(mode=context.GRAPH_MODE)
class MockNeg(PrimitiveWithInfer):
@prim_attr_register
def __init__(self):
@ -47,6 +46,7 @@ class MockNeg(PrimitiveWithInfer):
raise TypeError("InferError")
return input_x
class MockSub(PrimitiveWithInfer):
@prim_attr_register
def __init__(self):
@ -59,6 +59,7 @@ class MockSub(PrimitiveWithInfer):
def infer_dtype(self, x_dtype, y_dtype):
return x_dtype
@bprop_getters.register(MockSub)
def get_bprop_mock_sub(self):
"""Grad definition for `MockSub` operation."""
@ -66,8 +67,10 @@ def get_bprop_mock_sub(self):
def bprop(x, y, out, dout):
return binop_grad_common(x, y, dout, neg_func(dout))
return bprop
class Net(nn.Cell):
def __init__(self, in_features, out_features):
super(Net, self).__init__()
@ -80,6 +83,7 @@ class Net(nn.Cell):
output = self.add(self.matmul(input, self.weight), self.bias)
return output
class NetFP16(nn.Cell):
def __init__(self, in_features, out_features):
super(NetFP16, self).__init__()
@ -90,16 +94,19 @@ class NetFP16(nn.Cell):
self.cast = P.Cast()
def construct(self, input):
output = self.cast(self.add(self.matmul(self.cast(input, mstype.float16), self.cast(self.weight, mstype.float16)),
self.cast(self.bias, mstype.float16)), mstype.float32)
output = self.cast(
self.add(self.matmul(self.cast(input, mstype.float16), self.cast(self.weight, mstype.float16)),
self.cast(self.bias, mstype.float16)), mstype.float32)
return output
def get_axis(x):
shape = F.shape(x)
length = F.tuple_len(shape)
perm = F.make_range(0, length)
return perm
class MSELoss(nn.Cell):
def __init__(self):
super(MSELoss, self).__init__()
@ -107,17 +114,21 @@ class MSELoss(nn.Cell):
self.square = P.Square()
self.reduce_mean = P.ReduceMean()
self.sub = MockSub()
def construct(self, data, label):
diff = self.sub(data, label)
return self.reduce_mean(self.square(diff), get_axis(diff))
class NegCell(nn.Cell):
def __init__(self):
super(NegCell, self).__init__()
self.neg = MockNeg()
def construct(self, x):
return self.neg(x)
class Net3(nn.Cell):
def __init__(self):
super().__init__()
@ -146,6 +157,7 @@ class SequenceNet(nn.Cell):
x = self.seq(x) + bbb
return x
def test_sequential_resolve_error():
input_np = np.random.randn(2, 3, 4, 5).astype(np.float32)
input_me = Tensor(input_np)
@ -153,6 +165,7 @@ def test_sequential_resolve_error():
with pytest.raises(RuntimeError) as e:
net(input_me)
def test_compile_grad_error():
inputs = Tensor(np.ones([16, 16]).astype(np.float32))
label = Tensor(np.zeros([16, 16]).astype(np.float32))
@ -164,9 +177,8 @@ def test_compile_grad_error():
net_with_loss = WithLossCell(net, loss)
scale_manager = DynamicLossScaleManager()
update_cell = scale_manager.get_update_cell()
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell = update_cell)
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell=update_cell)
train_network.set_train()
with pytest.raises(TypeError) as e:
train_network(inputs, label)
print (e)
print(e)

41
tests/ut/python/optimizer/test_optimize_with_loss_scale.py
View File

@ -13,19 +13,20 @@
# limitations under the License.
# ============================================================================
import numpy as np
import mindspore.nn as nn
from mindspore import context
from mindspore import Tensor, Parameter
from mindspore import context
from mindspore.common import dtype as mstype
from mindspore.nn.optim import Lamb
from mindspore.nn.optim import Momentum, Adam
from mindspore.nn.wrap.cell_wrapper import WithLossCell
from mindspore.nn.wrap.loss_scale import TrainOneStepWithLossScaleCell
from mindspore.train.loss_scale_manager import FixedLossScaleManager, DynamicLossScaleManager
from mindspore.ops import operations as P
from mindspore.nn.optim import Momentum, Adam
from mindspore.ops import functional as F
from mindspore.common import dtype as mstype
from mindspore.ops import operations as P
from mindspore.train import Model
from mindspore.train.loss_scale_manager import FixedLossScaleManager, DynamicLossScaleManager
from ....dataset_mock import MindData
from mindspore.nn.optim import Lamb
context.set_context(mode=context.GRAPH_MODE)
@ -36,6 +37,7 @@ class MindDataSet(MindData):
np_types=dataset_types,
output_shapes=dataset_shapes,
input_indexs=(0, 1))
def __next__(self):
if self._size < self._iter_num:
raise StopIteration
@ -45,6 +47,7 @@ class MindDataSet(MindData):
next.append(Tensor(np.ones(shape).astype(type)))
return tuple(next)
class Net(nn.Cell):
def __init__(self, in_features, out_features):
super(Net, self).__init__()
@ -57,6 +60,7 @@ class Net(nn.Cell):
output = self.add(self.matmul(input, self.weight), self.bias)
return output
class NetFP16(nn.Cell):
def __init__(self, in_features, out_features):
super(NetFP16, self).__init__()
@ -67,10 +71,12 @@ class NetFP16(nn.Cell):
self.cast = P.Cast()
def construct(self, input):
output = self.cast(self.add(self.matmul(self.cast(input, mstype.float16), self.cast(self.weight, mstype.float16)),
self.cast(self.bias, mstype.float16)), mstype.float32)
output = self.cast(
self.add(self.matmul(self.cast(input, mstype.float16), self.cast(self.weight, mstype.float16)),
self.cast(self.bias, mstype.float16)), mstype.float32)
return output
def get_axis(x):
shape_op = P.Shape()
shape = shape_op(x)
@ -78,6 +84,7 @@ def get_axis(x):
perm = F.make_range(0, length)
return perm
class MSELoss(nn.Cell):
def __init__(self):
super(MSELoss, self).__init__()
@ -89,6 +96,7 @@ class MSELoss(nn.Cell):
diff = data - label
return self.reduce_mean(self.square(diff), get_axis(diff))
def test_momentum_compile():
inputs = Tensor(np.ones([15, 1]).astype(np.float32))
label = Tensor(np.zeros([15, 1]).astype(np.float32))
@ -104,6 +112,7 @@ def test_momentum_compile():
output = train_network(inputs, label, scaling_sens)
print("the result is ", output)
def test_compile_fp16_not_overflow():
inputs = Tensor(np.ones([16, 16]).astype(np.float32))
label = Tensor(np.zeros([16, 16]).astype(np.float32))
@ -119,6 +128,7 @@ def test_compile_fp16_not_overflow():
output = train_network(inputs, label, scaling_sens)
print("the result is ", output)
def test_compile_fp16_lr_overflow():
inputs = Tensor(np.ones([16, 16]).astype(np.float32))
label = Tensor(np.zeros([16, 16]).astype(np.float32))
@ -134,6 +144,7 @@ def test_compile_fp16_lr_overflow():
output = train_network(inputs, label, scaling_sens)
print("the result is ", output)
def test_compile_fp16_overflow():
inputs = Tensor(np.ones([16, 16]).astype(np.float32))
label = Tensor(np.zeros([16, 16]).astype(np.float32))
@ -148,6 +159,7 @@ def test_compile_fp16_overflow():
output = train_network(inputs, label, scaling_sens)
print("the result is ", output)
def test_compile_fp16_lr_overflow_with_lossscale_update():
inputs = Tensor(np.ones([16, 16]).astype(np.float32))
label = Tensor(np.zeros([16, 16]).astype(np.float32))
@ -165,6 +177,7 @@ def test_compile_fp16_lr_overflow_with_lossscale_update():
output = train_network(inputs, label, scaling_sens)
print("the result is ", output)
def test_compile_f16_model_train():
dataset_types = (np.float32, np.float32)
dataset_shapes = ((16, 16), (16, 16))
@ -205,11 +218,12 @@ def test_compile_fp16_lr_overflow_fixed_feed():
net_with_loss = WithLossCell(net, loss)
scale_manager = FixedLossScaleManager()
update_cell = scale_manager.get_update_cell()
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell = update_cell)
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell=update_cell)
train_network.set_train()
output = train_network(inputs, label, scaling_sens)
print("the result is ", output)
def test_compile_fp16_lr_overflow_dynamic_feed():
inputs = Tensor(np.ones([16, 16]).astype(np.float32))
label = Tensor(np.zeros([16, 16]).astype(np.float32))
@ -222,11 +236,12 @@ def test_compile_fp16_lr_overflow_dynamic_feed():
net_with_loss = WithLossCell(net, loss)
scale_manager = DynamicLossScaleManager()
update_cell = scale_manager.get_update_cell()
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell = update_cell)
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell=update_cell)
train_network.set_train()
output = train_network(inputs, label, scaling_sens)
print("the result is ", output)
def test_compile_fp16_lr_overflow_fixed_graph():
inputs = Tensor(np.ones([16, 16]).astype(np.float32))
label = Tensor(np.zeros([16, 16]).astype(np.float32))
@ -238,11 +253,12 @@ def test_compile_fp16_lr_overflow_fixed_graph():
net_with_loss = WithLossCell(net, loss)
scale_manager = FixedLossScaleManager(drop_overflow_update=True)
update_cell = scale_manager.get_update_cell()
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell = update_cell)
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell=update_cell)
train_network.set_train()
output = train_network(inputs, label)
print("the result is ", output)
def test_compile_fp16_lr_overflow_dynamic_graph():
inputs = Tensor(np.ones([16, 16]).astype(np.float32))
label = Tensor(np.zeros([16, 16]).astype(np.float32))
@ -254,11 +270,12 @@ def test_compile_fp16_lr_overflow_dynamic_graph():
net_with_loss = WithLossCell(net, loss)
scale_manager = DynamicLossScaleManager()
update_cell = scale_manager.get_update_cell()
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell = update_cell)
train_network = TrainOneStepWithLossScaleCell(net_with_loss, optimizer, scale_update_cell=update_cell)
train_network.set_train()
output = train_network(inputs, label)
print("the result is ", output)
def test_adam_compile():
inputs = Tensor(np.ones([15, 1]).astype(np.float32))
label = Tensor(np.zeros([15, 1]).astype(np.float32))

6
tests/ut/python/optimizer/test_optimize_with_parameter_groups.py
View File

@ -14,20 +14,22 @@
# ============================================================================
import numpy as np
import pytest
import mindspore.common.dtype as mstype
import mindspore.nn as nn
from mindspore.nn.optim import Momentum, SGD, RMSProp, Adam
from mindspore import context
from mindspore.common.api import _executor
from mindspore.common.tensor import Tensor
from mindspore.ops import operations as P
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.nn.optim import Momentum, SGD, RMSProp, Adam
from mindspore.ops import operations as P
context.set_context(mode=context.GRAPH_MODE)
class LeNet5(nn.Cell):
""" LeNet5 definition """
def __init__(self):
super(LeNet5, self).__init__()
self.conv1 = nn.Conv2d(1, 6, 5, pad_mode='valid')

24
tests/ut/python/parameter_feature/test_parameter.py
View File

@ -13,16 +13,17 @@
# limitations under the License.
# ============================================================================
import numpy as np
import mindspore.context as context
import mindspore.ops.composite as C
from mindspore import Tensor, Parameter
from mindspore.nn import Cell
from mindspore.ops import operations as P
import mindspore.ops.composite as C
context.set_context(mode=context.GRAPH_MODE, save_graphs=True)
def test_parser_three_default_mixed_args_subnet():
def test_parser_three_default_mixed_args_subnet():
class SubNetDefaultMixedArgs(Cell):
def __init__(self):
super().__init__()
@ -55,7 +56,7 @@ def test_net_vararg_kwonlyarg_kwarg():
super(FirstNet, self).__init__()
self.net = SecondNet()
def construct(self, x=1, z=2+2+4, y=3):
def construct(self, x=1, z=2 + 2 + 4, y=3):
c = self.net(22, 33, x, y, z, 2, 3, 4, 5, key1=10, key2=20, key3=30, key4=40)
return c
@ -74,13 +75,14 @@ def test_net_vararg_kwonlyarg_kwarg():
net = FirstNet()
net()
def test_net_vararg_normal_input():
class FirstNet(Cell):
def __init__(self):
super(FirstNet, self).__init__()
self.net = SecondNet()
def construct(self, x=1, z=2+2+4, y=3):
def construct(self, x=1, z=2 + 2 + 4, y=3):
c = self.net(22, 33, x, y, z, 2, 3, 4, 5, key1=10, key2=20, key3=30, key4=40)
return c
@ -95,10 +97,12 @@ def test_net_vararg_normal_input():
d = var[0] * var[1] * var[2] * var[3]
e = key1 - key2 - kwargs["key3"] + kwargs["key4"]
return a + b + c + d + e
x = Tensor(np.ones((2, 3, 4), np.int32))
net = FirstNet()
net(x, x, x)
def test_prim_vararg_kwonlyarg():
class FirstNet(Cell):
def __init__(self):
@ -201,9 +205,11 @@ def test_net_variable_and_weights():
z = Tensor(np.ones((4,), np.float32))
net(x, y, z)
def test_net_vargs_expand():
class InputBackward(Cell):
""" InputBackward definition """
def __init__(self, network, c1=None, c2=None):
super(InputBackward, self).__init__()
self.network = network
@ -214,9 +220,11 @@ def test_net_vargs_expand():
def construct(self, *inputs):
return self.grad(self.network)(*inputs)
class AddNet(Cell):
def __init__(self):
super(AddNet, self).__init__()
def construct(self, x, y):
return x + y
@ -227,7 +235,7 @@ def test_net_vargs_expand():
net.set_train()
net(x, y, sens)
def test_mixed_precision_const_parameter():
class NetLoss(Cell):
@ -237,6 +245,7 @@ def test_mixed_precision_const_parameter():
self.up_sample1 = P.ResizeBilinear((14, 14))
self.up_sample2 = P.ResizeBilinear((28, 28))
self.up_sample3 = P.ResizeBilinear((36, 36))
def construct(self, x, y, z, *args):
ret = 0
if args[0] == self.shape(z)[2]:
@ -250,20 +259,23 @@ def test_mixed_precision_const_parameter():
ret = x * y
ret = ret * z
return ret
class NetMain(Cell):
def __init__(self, loss_fn):
super(NetMain, self).__init__()
self.loss_fn = loss_fn
self.shape = P.Shape()
def construct(self, x, y, z):
size_x = self.shape(x)[2]
size_y = self.shape(y)[2]
ret = self.loss_fn(x, y, z, size_x, size_y)
return ret
loss_fn = NetLoss()
net = NetMain(loss_fn)
net.add_flags_recursive(fp32=True)
x = Tensor(np.ones((1, 3, 28, 28), np.float32))
y = Tensor(np.ones((1, 3, 14, 14), np.float32))
z = Tensor(np.ones((1, 3, 28, 28), np.float32))
out = net(x, y, z)
out = net(x, y, z)

26
tests/ut/python/parameter_feature/test_var_grad.py
View File

@ -13,14 +13,14 @@
# limitations under the License.
# ============================================================================
import numpy as np
from mindspore import context
import mindspore.ops.composite as C
from mindspore import Tensor, Parameter
from mindspore import context
from mindspore.common import dtype as mstype
from mindspore.common.parameter import ParameterTuple
from mindspore.nn import Cell
from mindspore.ops import operations as P
import mindspore.ops.composite as C
from mindspore.common.api import _executor
from mindspore.common.parameter import ParameterTuple
from mindspore.common import dtype as mstype
context.set_context(mode=context.GRAPH_MODE)
@ -34,6 +34,7 @@ def test_net_vargs_expand():
def construct(self, x, y):
return x + y
x = Tensor(np.random.normal(0, 1, [3, 4, 5]).astype(np.float32))
y = Tensor(np.random.normal(0, 1, [3, 4, 5]).astype(np.float32))
sens = Tensor(np.random.normal(0, 1, [3, 4, 5]).astype(np.float32))
@ -51,7 +52,7 @@ class VarNet(Cell):
self.net = net
def construct(self, *args):
return self.net(*args)*self.w + self.b
return self.net(*args) * self.w + self.b
class SecondNet(Cell):
@ -95,6 +96,7 @@ class Bprop(Cell):
def test_all_var_args_grad_with_sens():
""""test grad_by_list_with_sens with all var args input"""
class GradNet(Cell):
def __init__(self, net):
super(GradNet, self).__init__()
@ -103,6 +105,7 @@ def test_all_var_args_grad_with_sens():
def construct(self, *inputs):
return C.grad_by_list_with_sens(self.net, self.weights)(*inputs)
x = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
y = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
sens = Tensor(1.0, dtype=mstype.float32)
@ -120,6 +123,7 @@ def test_grad_list_var_args():
def construct(self, *inputs):
return C.grad_by_list(self.net, self.weights)(*inputs)
x = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
y = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
net = VarNet(SecondNet())
@ -136,6 +140,7 @@ def test_grad_all_var_args():
def construct(self, *inputs):
return C.grad_all(self.net)(*inputs)
x = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
y = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
net = VarNet(SecondNet())
@ -152,6 +157,7 @@ def test_grad_all_var_args_with_sens():
def construct(self, *inputs):
return C.grad_all_with_sens(self.net)(*inputs)
x = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
y = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
sens = Tensor(1.0, dtype=mstype.float32)
@ -169,6 +175,7 @@ def test_grad_var_args_with_sens():
def construct(self, *inputs):
return C.grad_with_sens(self.net)(*inputs)
x = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
y = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
sens = Tensor(1.0, dtype=mstype.float32)
@ -206,6 +213,7 @@ def test_var_args_grad():
def construct(self, x, y, sens):
return C.grad_by_list_with_sens(self.net, self.weights)(x, y, sens)
x = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
y = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
sens = Tensor(1.0, dtype=mstype.float32)
@ -216,13 +224,14 @@ def test_var_args_grad():
def test_var_args_positional():
""""test grad_all with var args in inner graph"""
class VarNet(Cell):
def __init__(self, net):
super(VarNet, self).__init__()
self.net = net
def construct(self, x, y):
return self.net(x, y)*x
return self.net(x, y) * x
class SecondNet(Cell):
def __init__(self):
@ -239,6 +248,7 @@ def test_var_args_positional():
def construct(self, x, y):
return C.grad_all(self.net)(x, y)
x = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
y = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
net = VarNet(SecondNet())
@ -258,6 +268,7 @@ def test_grad_within_if_else():
def construct(self, *inputs):
return self.grad(*inputs)
x = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
y = Tensor(np.ones([3, 4, 5]), dtype=mstype.float32)
sens = Tensor(1.0, dtype=mstype.float32)
@ -309,6 +320,7 @@ def test_grad_for_concat():
def grad_cmp(self):
input_grad_mindspore = self.grad_mindspore_impl()
fact = ConcatFactory(input_shape=(
(2, 184320, 1), (2, 46080, 1), (2, 11520, 1), (2, 2880, 1), (2, 720, 1)), axis=1)
fact.grad_cmp()

22
tests/ut/python/test_log.py
View File

@ -15,12 +15,13 @@
"""
log test
"""
import logging
import os
import sys
import time
import re
import shutil
import logging
import sys
import time
def test_log_stdout():
# Clean up environment variables
@ -50,8 +51,8 @@ def test_log_setlevel():
_rm_env_config()
os.environ['GLOG_v'] = '0'
from mindspore import log as logger
#logger_instance = logger._get_logger()
#del logger_instance
# logger_instance = logger._get_logger()
# del logger_instance
loglevel = logger.get_level()
log_str = 'print debug informations'
logger.debug("5 test log message debug:%s", log_str)
@ -87,7 +88,7 @@ def test_log_file():
"\[.*:.*[0-9]\] test log message warning"
match_obj = re.match(pattern, result)
#Clear test file
# Clear test file
if os.path.exists(file_path):
shutil.rmtree(file_path)
@ -100,7 +101,7 @@ def test_log_backup_count():
"""
test backup count
"""
#logger.reset_log_config(level=logging.INFO, console=False,
# logger.reset_log_config(level=logging.INFO, console=False,
# filepath=file_path, maxBytes=1000, backupCount=10)
_rm_env_config()
file_path = '/tmp/log/mindspore_test'
@ -141,12 +142,12 @@ def test_log_verify_envconfig():
# level is not a number
_rm_env_config()
os.environ['GLOG_v'] = 'test'
verify_dict_0 = logger._get_env_config()
verify_dict_0 = logger._get_env_config()
# level is not in range
_rm_env_config()
os.environ['GLOG_v'] = '100'
verify_dict_1 = logger._get_env_config()
verify_dict_1 = logger._get_env_config()
# console is not a number
_rm_env_config()
@ -236,7 +237,6 @@ def test_log_repeated_print():
logger._global_logger = None
def test_log_getconfig():
_rm_env_config()
os.environ['GLOG_v'] = '3'
@ -307,7 +307,7 @@ def test_log_ms_import():
targetdict = {'GLOG_v': '2', 'GLOG_logtostderr': '1'}
level = ms.get_level()
assert configdict == targetdict and level == '2'
def _rm_env_config():
envlist = ['GLOG_v', 'GLOG_logtostderr', 'GLOG_log_dir', 'logger_maxBytes', 'logger_backupCount']

1
tests/ut/python/train/summary/__init__.py
View File

@ -15,6 +15,7 @@
"""setup for pytest"""
import mindspore.context as context
# pylint: disable=unused-argument
def setup_module(module):
context.set_context(mode=context.GRAPH_MODE)

9
tests/ut/python/train/summary/test_graph_summary.py
View File

@ -13,15 +13,16 @@
# limitations under the License.
# ============================================================================
""" test_graph_summary """
import os
import logging
import os
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore.nn.optim import Momentum
from mindspore import Model, context
from mindspore.train.summary.summary_record import SummaryRecord
from mindspore.nn.optim import Momentum
from mindspore.train.callback import SummaryStep
from mindspore.train.summary.summary_record import SummaryRecord
from .....dataset_mock import MindData
CUR_DIR = os.getcwd()

2
tests/ut/python/train/summary/test_histogram_summary.py
View File

@ -21,8 +21,8 @@ import tempfile
import numpy as np
from mindspore.common.tensor import Tensor
from mindspore.train.summary.summary_record import SummaryRecord, _cache_summary_tensor_data
from mindspore.train.summary._summary_adapter import _calc_histogram_bins
from mindspore.train.summary.summary_record import SummaryRecord, _cache_summary_tensor_data
from .summary_reader import SummaryReader
CUR_DIR = os.getcwd()

10
tests/ut/python/train/summary/test_image_summary.py
View File

@ -18,16 +18,18 @@
@Date : 2019-07-4
@Desc : test summary function
"""
import os
import logging
import os
import numpy as np
import mindspore.nn as nn
from mindspore.train.summary.summary_record import SummaryRecord, \
_cache_summary_tensor_data
from mindspore import Model, context
from mindspore import Tensor
from mindspore.nn.optim import Momentum
from mindspore import Model, context
from mindspore.train.callback import SummaryStep
from mindspore.train.summary.summary_record import SummaryRecord, \
_cache_summary_tensor_data
from .....dataset_mock import MindData
CUR_DIR = os.getcwd()

21
tests/ut/python/train/summary/test_summary.py
View File

@ -18,16 +18,18 @@
@Date : 2019-07-4
@Desc : test summary function
"""
import os
import logging
import os
import random
import numpy as np
import pytest
from mindspore.train.summary.summary_record import SummaryRecord, _cache_summary_tensor_data
from mindspore.train.callback import SummaryStep
from mindspore.common.tensor import Tensor
import mindspore.nn as nn
from mindspore.common.tensor import Tensor
from mindspore.ops import operations as P
from mindspore.train.callback import SummaryStep
from mindspore.train.summary.summary_record import SummaryRecord, _cache_summary_tensor_data
CUR_DIR = os.getcwd()
SUMMARY_DIR = CUR_DIR + "/test_temp_summary_event_file/"
@ -129,7 +131,8 @@ def test_scalar_summary_sample_with_shape_1():
# Test: test with ge
class SummaryDemo(nn.Cell):
""" SummaryDemo definition """
def __init__(self,):
def __init__(self, ):
super(SummaryDemo, self).__init__()
self.s = P.ScalarSummary()
self.histogram_summary = P.HistogramSummary()
@ -218,9 +221,9 @@ def test_validate():
with pytest.raises(ValueError):
sr.record(2.0)
with pytest.raises(ValueError):
sr.record((1,3))
sr.record((1, 3))
with pytest.raises(ValueError):
sr.record([2,3])
sr.record([2, 3])
with pytest.raises(ValueError):
sr.record("str")
with pytest.raises(ValueError):
@ -235,8 +238,8 @@ def test_validate():
with pytest.raises(ValueError):
SummaryStep(sr, "str")
with pytest.raises(ValueError):
SummaryStep(sr, (1,2))
SummaryStep(sr, (1, 2))
with pytest.raises(ValueError):
SummaryStep(sr, [3,4])
SummaryStep(sr, [3, 4])
with pytest.raises(ValueError):
SummaryStep(sr, sr)

8
tests/ut/python/train/summary/test_summary_abnormal_input.py
View File

@ -18,11 +18,13 @@
@Date : 2019-08-5
@Desc : test summary function of abnormal input
"""
import os
import logging
import os
import numpy as np
from mindspore.train.summary.summary_record import SummaryRecord
from mindspore.common.tensor import Tensor
from mindspore.train.summary.summary_record import SummaryRecord
CUR_DIR = os.getcwd()
SUMMARY_DIR = CUR_DIR + "/test_temp_summary_event_file/"
@ -65,6 +67,7 @@ def test_summaryrecord_input_null_string():
assert False
log.debug("finished test_summaryrecord_input_null_string")
def test_summaryrecord_input_None():
log.debug("begin test_summaryrecord_input_None")
# step 0: create the thread
@ -76,6 +79,7 @@ def test_summaryrecord_input_None():
assert False
log.debug("finished test_summaryrecord_input_None")
def test_summaryrecord_input_relative_dir_1():
log.debug("begin test_summaryrecord_input_relative_dir_1")
# step 0: create the thread

21
tests/ut/python/train/summary/test_summary_ops_params_valid_check.py
View File

@ -18,16 +18,17 @@
@Date : 2019-08-5
@Desc : test summary function of ops params valid check
"""
import os
import logging
import os
import random
import numpy as np
import pytest
from mindspore.train.summary.summary_record import SummaryRecord
from mindspore.common.tensor import Tensor
import mindspore.nn as nn
from mindspore.ops import operations as P
import mindspore.nn as nn
from mindspore.common.tensor import Tensor
from mindspore.ops import operations as P
from mindspore.train.summary.summary_record import SummaryRecord
CUR_DIR = os.getcwd()
SUMMARY_DIR = CUR_DIR + "/test_temp_summary_event_file/"
@ -38,6 +39,7 @@ log.setLevel(level=logging.ERROR)
class SummaryDemoTag(nn.Cell):
""" SummaryDemoTag definition """
def __init__(self, tag1, tag2, tag3):
super(SummaryDemoTag, self).__init__()
self.s = P.ScalarSummary()
@ -58,6 +60,7 @@ class SummaryDemoTag(nn.Cell):
class SummaryDemoTagForSet(nn.Cell):
""" SummaryDemoTagForSet definition """
def __init__(self, tag_tuple):
super(SummaryDemoTagForSet, self).__init__()
self.s = P.ScalarSummary()
@ -75,6 +78,7 @@ class SummaryDemoTagForSet(nn.Cell):
class SummaryDemoValue(nn.Cell):
""" SummaryDemoValue definition """
def __init__(self, value):
super(SummaryDemoValue, self).__init__()
self.s = P.ScalarSummary()
@ -88,8 +92,10 @@ class SummaryDemoValue(nn.Cell):
self.s("y", self.v)
return z
class SummaryDemoValueForSet(nn.Cell):
""" SummaryDemoValueForSet definition """
def __init__(self, value, tag_tuple):
super(SummaryDemoValueForSet, self).__init__()
self.s = P.ScalarSummary()
@ -106,11 +112,12 @@ class SummaryDemoValueForSet(nn.Cell):
class HistogramSummaryNet(nn.Cell):
"HistogramSummaryNet definition"
def __init__(self, value):
self.histogram_summary = P.HistogramSummary()
self.add = P.TensorAdd()
self.value = value
def construct(self, tensors1, tensor2):
self.histogram_summary("value", self.value)
return self.add(tensors1, tensor2)
@ -246,7 +253,7 @@ def test_histogram_summary_use_valid_value():
"""Test histogram summary with valid value"""
log.debug("Begin test_histogram_summary_use_valid_value")
try:
net = HistogramSummaryNet(Tensor(np.array([1,2,3])))
net = HistogramSummaryNet(Tensor(np.array([1, 2, 3])))
run_case(net)
except:
assert True

15
tests/ut/python/train/summary/test_tensor_summary.py
View File

@ -18,13 +18,15 @@
@Date : 2019-07-4
@Desc : test summary function
"""
import os
import logging
import os
import numpy as np
from mindspore.train.summary.summary_record import SummaryRecord, _cache_summary_tensor_data
from mindspore.common.tensor import Tensor
import mindspore.nn as nn
from mindspore.common.tensor import Tensor
from mindspore.ops import operations as P
from mindspore.train.summary.summary_record import SummaryRecord, _cache_summary_tensor_data
CUR_DIR = os.getcwd()
SUMMARY_DIR = CUR_DIR + "/test_temp_summary_event_file/"
@ -93,7 +95,6 @@ def test_tensor_summary_sample():
log.debug("finished test_tensor_summary_sample")
def get_test_data_check(step):
""" get_test_data_check """
test_data_list = []
@ -111,7 +112,8 @@ def get_test_data_check(step):
# Test: test with ge
class SummaryDemo(nn.Cell):
""" SummaryDemo definition """
def __init__(self,):
def __init__(self, ):
super(SummaryDemo, self).__init__()
self.s = P.TensorSummary()
self.add = P.TensorAdd()
@ -123,6 +125,7 @@ class SummaryDemo(nn.Cell):
self.s("y1", y)
return z
def test_tensor_summary_with_ge():
""" test_tensor_summary_with_ge """
log.debug("begin test_tensor_summary_with_ge")
@ -140,7 +143,7 @@ def test_tensor_summary_with_ge():
steps = 100
for i in range(1, steps):
x = Tensor(np.array([[i], [i]]).astype(np.float32))
y = Tensor(np.array([[i+1], [i+1]]).astype(np.float32))
y = Tensor(np.array([[i + 1], [i + 1]]).astype(np.float32))
net(x, y)
test_writer.record(i)

9
tests/ut/python/train/test_amp.py
View File

@ -15,12 +15,11 @@
""" auto mixed precision """
import numpy as np
import pytest
import mindspore.context as context
from mindspore import Tensor
from mindspore import amp
from mindspore import nn
from mindspore import Tensor
from mindspore.common import dtype as mstype
import mindspore.context as context
from mindspore.model_zoo.resnet import resnet50
from mindspore.train import Model
from ....dataset_mock import MindData
@ -96,6 +95,7 @@ class MindDataSet(MindData):
np_types=dataset_types,
output_shapes=dataset_shapes,
input_indexs=(0, 1))
def __next__(self):
if self._size < self._iter_num:
raise StopIteration
@ -122,6 +122,7 @@ def test_compile_model_train_O0():
# not actual run, the metrics step will fail, check if compile ok.
model.eval(dataset)
def test_compile_model_train_O2():
dataset_types = (np.float32, np.float32)
dataset_shapes = ((16, 16), (16, 16))

1
tests/ut/python/train/test_run_config.py
View File

@ -14,6 +14,7 @@
# ============================================================================
""" test_run_config """
import pytest
from mindspore.train.callback import CheckpointConfig

4
tests/ut/python/train/test_training.py
View File

@ -14,12 +14,14 @@
# ============================================================================
""" test_training """
import logging
import numpy as np
import pytest
import mindspore.nn as nn
from mindspore import Model, context
from mindspore import Tensor
from mindspore.nn.optim import Momentum
from mindspore import Model, context
from mindspore.train.callback import SummaryStep
from ..ut_filter import non_graph_engine
from ....dataset_mock import MindData

13
tests/ut/python/transform/test_transform.py
View File

@ -19,12 +19,12 @@
@Desc : test mindspore compile method
"""
import logging
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor, Parameter, Model
from mindspore.ops import operations as P
from ..ut_filter import non_graph_engine
import numpy as np
import mindspore.nn as nn
from mindspore import Tensor, Parameter
from mindspore.ops import operations as P
log = logging.getLogger("test")
log.setLevel(level=logging.ERROR)
@ -104,6 +104,7 @@ class ResidualBlock(nn.Cell):
class ResNet(nn.Cell):
""" ResNet definition """
def __init__(self, tensor):
super(ResNet, self).__init__()
self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3)
@ -118,6 +119,7 @@ class ResNet(nn.Cell):
class LeNet(nn.Cell):
""" LeNet definition """
def __init__(self):
super(LeNet, self).__init__()
self.relu = nn.ReLU()
@ -165,4 +167,3 @@ class Net(nn.Cell):
def construct(self, input_x):
return self.softmax(input_x)

8
tests/ut/python/utils/test_callback.py
View File

@ -15,19 +15,19 @@
"""test callback function."""
import os
import stat
import numpy as np
import pytest
import mindspore.nn as nn
import mindspore.common.dtype as mstype
from mindspore import context
import mindspore.nn as nn
from mindspore.common.api import ms_function
from mindspore.common.tensor import Tensor
from mindspore.nn.optim import Momentum
from mindspore.nn import TrainOneStepCell, WithLossCell
from mindspore.nn.optim import Momentum
from mindspore.train.callback import ModelCheckpoint, _check_file_name_prefix, RunContext, _checkpoint_cb_for_save_op, \
LossMonitor, _InternalCallbackParam, _chg_ckpt_file_name_if_same_exist, \
_build_callbacks, CheckpointConfig, _set_cur_net
from mindspore.common.api import ms_function
class Net(nn.Cell):

2
tests/ut/python/utils/test_checkparam.py
View File

@ -15,6 +15,7 @@
""" test_checkparam """
import numpy as np
import pytest
import mindspore
import mindspore.nn as nn
from mindspore import Model, context
@ -23,6 +24,7 @@ from mindspore.common.tensor import Tensor
class LeNet5(nn.Cell):
""" LeNet5 definition """
def __init__(self):
super(LeNet5, self).__init__()
self.conv1 = nn.Conv2d(3, 6, 5, pad_mode="valid")

Some files were not shown because too many files have changed in this diff Show More