mindspore/tests/ut/python/fallback/test_graph_fallback_parse.py

259 lines
6.4 KiB
Python
Raw Normal View History

# Copyright 2022 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ============================================================================
""" test graph fallback """
import pytest
import numpy as np
import mindspore.nn as nn
from mindspore import context, Tensor
context.set_context(mode=context.GRAPH_MODE)
def test_parse_return():
"""
Feature: JIT Fallback
Description: Test Interpret node in return statement in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def __init__(self):
super(Network, self).__init__()
self.x = np.array([1, 2, 3])
def construct(self):
return Tensor(self.x)
net = Network()
out = net()
assert (out.asnumpy() == [1, 2, 3]).all()
def test_parse_def_function():
"""
Feature: JIT Fallback
Description: Test args default value is Interpret node in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def __init__(self):
super(Network, self).__init__()
self.x = np.array([1, 2, 3])
def construct(self, y=np.array([2, 3, 4])):
return Tensor(self.x + y)
net = Network()
out = net()
assert (out.asnumpy() == [3, 5, 7]).all()
def test_parse_lambda():
"""
Feature: JIT Fallback
Description: Test Interpret node in lambda in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def __init__(self):
super(Network, self).__init__()
self.x = np.array([1, 2, 3])
def construct(self):
new_x = lambda x: 2 * x + self.x
y = new_x(1)
return Tensor(y)
net = Network()
out = net()
assert (out.asnumpy() == [3, 4, 5]).all()
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_parse_lambda_2():
"""
Feature: JIT Fallback
Description: Test Interpret node in lambda in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def __init__(self):
super(Network, self).__init__()
self.x = np.array([1, 2, 3])
def construct(self):
new_x = lambda x: 2 * x + self.x
return Tensor(new_x(1))
net = Network()
out = net()
assert (out.asnumpy() == [3, 4, 5]).all()
def test_parse_bool_op():
"""
Feature: JIT Fallback
Description: Test Interpret node in bool op in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def __init__(self):
super(Network, self).__init__()
self.x = np.array([1, 2, 3])
def construct(self):
if Tensor(2) and self.x.all():
return Tensor(self.x + 1)
return Tensor(self.x)
net = Network()
out = net()
assert (out.asnumpy() == [2, 3, 4]).all()
def test_parse_tuple():
"""
Feature: JIT Fallback
Description: Test Interpret node in tuple in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def construct(self):
x = Tensor([1])
tuple_num = [x, x + 1, Tensor([3])]
return tuple_num
net = Network()
out = net()
assert out[0].asnumpy() == 1 and out[1].asnumpy() == 2 and out[2].asnumpy() == 3
def test_parse_slice():
"""
Feature: JIT Fallback
Description: Test Interpret node in slice in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def construct(self):
x = [Tensor([11]), Tensor([22]), Tensor([33])]
y = x[Tensor([0]): Tensor([2])]
return y
net = Network()
out = net()
assert out[0].asnumpy() == 11 and out[1].asnumpy() == 22
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_parse_subscript():
"""
Feature: JIT Fallback
Description: Test Interpret node in subscript in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def construct(self):
x = [Tensor([11]), Tensor([22]), Tensor([33])]
y = x[Tensor([0])] + x[Tensor([1])] + x[Tensor([2])]
return y
net = Network()
out = net()
assert out.asnumpy() == 66
def test_parse_subscript_2():
"""
Feature: JIT Fallback
Description: Test Interpret node in subscript in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def construct(self):
x = [Tensor([11]), Tensor([22]), Tensor([33])]
y = x[np.array(0)]
return y
net = Network()
out = net()
assert out.asnumpy() == 11
def test_parse_unary_op():
"""
Feature: JIT Fallback
Description: Test Interpret node in unary op in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def construct(self):
x = np.array([1, 2, 3])
y = -x
return Tensor(y)
net = Network()
out = net()
assert (out.asnumpy() == [-1, -2, -3]).all()
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_parse_dict():
"""
Feature: JIT Fallback
Description: Test Interpret node in dict in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def construct(self):
x = {"a": np.array([1, 2, 3]), "b": np.array([4, 5, 6])}
key = "b"
try:
value = x[key]
except KeyError:
print("The key is not exist in x.")
return Tensor(value)
net = Network()
out = net()
assert (out.asnumpy() == [4, 5, 6]).all()
def test_parse_ifexpr():
"""
Feature: JIT Fallback
Description: Test Interpret node in ifexpr in graph mode.
Expectation: No exception.
"""
class Network(nn.Cell):
def construct(self):
y = Tensor([0]) if np.array([1]) else Tensor([1])
return y
net = Network()
out = net()
assert out == 0