!26718 [ME][Fallback] Add some fallback numpy test cases

Merge pull request !26718 from Margaret_wangrui/fallback
This commit is contained in:
i-robot 2021-11-24 06:43:13 +00:00 committed by Gitee
commit 2cc51099b1
2 changed files with 225 additions and 193 deletions

View File

@ -0,0 +1,179 @@
# Copyright 2021 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ============================================================================
""" test graph fallback """
import pytest
import numpy as np
from mindspore import ms_function, context, Tensor
context.set_context(mode=context.GRAPH_MODE)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_linspace():
"""
Feature: JIT Fallback
Description: Test numpy with linspace in graph mode.
Expectation: No exception.
"""
@ms_function
def np_linspace():
a = Tensor(np.linspace(1, 10, 10))
b = Tensor(np.linspace(1, 1, 10))
c = Tensor(np.linspace(10, 20, 5, endpoint=False))
d = Tensor(np.linspace(10, 20, 5, endpoint=True))
e = Tensor(np.linspace(1, 10, 10, retstep=True))
f = Tensor(np.linspace(1, 10, 10).reshape([10, 1]))
return a, b, c, d, e, f
a, b, c, d, e, f = np_linspace()
print("a:", a)
print("b:", b)
print("c:", c)
print("d:", d)
print("e:", e)
print("f:", f)
@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.platform_arm_ascend_training
@pytest.mark.platform_x86_ascend_training
@pytest.mark.env_onecard
def test_np_arange_slice_1():
"""
Feature: JIT Fallback
Description: Test numpy with arange slice in graph mode.
Expectation: No exception.
"""
@ms_function
def np_arange_slice_1():
x = np.arange(10)
index = slice(2, 7, 2)
a = Tensor(x[index])
b = Tensor(x[2:7:2])
c = Tensor(x[5])
d = Tensor(x[2:])
e = Tensor(x[2:5])
return a, b, c, d, e
a, b, c, d, e = np_arange_slice_1()
assert np.all(a.asnumpy() == Tensor(np.array([2, 4, 6])).asnumpy())
assert np.all(b.asnumpy() == Tensor(np.array([2, 4, 6])).asnumpy())
assert np.all(c.asnumpy() == Tensor(np.array([5])).asnumpy())
assert np.all(d.asnumpy() == Tensor(np.array([2, 3, 4, 5, 6, 7, 8, 9])).asnumpy())
assert np.all(e.asnumpy() == Tensor(np.array([2, 3, 4])).asnumpy())
@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.platform_arm_ascend_training
@pytest.mark.platform_x86_ascend_training
@pytest.mark.env_onecard
def test_np_arange_slice_2():
"""
Feature: JIT Fallback
Description: Test numpy with arange slice in graph mode.
Expectation: No exception.
"""
@ms_function
def np_arange_slice_2():
x = np.array([[1, 2, 3], [3, 4, 5], [4, 5, 6]])
a = Tensor(x[1:])
b = Tensor(x[..., 1])
c = Tensor(x[1, ...])
d = Tensor(x[..., 1:])
return a, b, c, d
a, b, c, d = np_arange_slice_2()
assert np.all(a.asnumpy() == Tensor(np.array([[3, 4, 5], [4, 5, 6]])).asnumpy())
assert np.all(b.asnumpy() == Tensor(np.array([2, 4, 5])).asnumpy())
assert np.all(c.asnumpy() == Tensor(np.array([3, 4, 5])).asnumpy())
assert np.all(d.asnumpy() == Tensor(np.array([[2, 3], [4, 5], [5, 6]])).asnumpy())
@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.platform_arm_ascend_training
@pytest.mark.platform_x86_ascend_training
@pytest.mark.env_onecard
def test_np_array_advanced_index_1():
"""
Feature: JIT Fallback
Description: Test numpy with array advanced index in graph mode.
Expectation: No exception.
"""
@ms_function
def np_array_advanced_index_1():
x = np.array([[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11]])
a = Tensor(x[[0, 1, 2], [0, 1, 0]])
rows = np.array([[0, 0], [3, 3]])
cols = np.array([[0, 2], [0, 2]])
b = Tensor(x[rows, cols])
c = Tensor(x[1:3, 1:3])
d = Tensor(x[1:3, [1, 2]])
e = Tensor(x[..., 1:])
return a, b, c, d, e
a, b, c, d, e = np_array_advanced_index_1()
assert np.all(a.asnumpy() == Tensor(np.array([0, 4, 6])).asnumpy())
assert np.all(b.asnumpy() == Tensor(np.array([[0, 2], [9, 11]])).asnumpy())
assert np.all(c.asnumpy() == Tensor(np.array([[4, 5], [7, 8]])).asnumpy())
assert np.all(d.asnumpy() == Tensor(np.array([[4, 5], [7, 8]])).asnumpy())
assert np.all(e.asnumpy() == Tensor(np.array([[1, 2], [4, 5], [7, 8], [10, 11]])).asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_array_advanced_index_2():
"""
Feature: JIT Fallback
Description: Test numpy with array advanced index in graph mode.
Expectation: No exception.
"""
@ms_function
def np_array_advanced_index_2():
x = np.array([[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11]])
y = np.array([np.nan, 1, 2, np.nan, 3, 4, 5])
z = np.array([1, 2 + 6j, 5, 3.5 + 5j])
a = Tensor(x[x > 5])
b = Tensor(y[~np.isnan(y)])
c = Tensor(z[np.iscomplex(z)])
return a, b, c
a, b, c = np_array_advanced_index_2()
assert np.all(a.asnumpy() == Tensor(np.array([6, 7, 8, 9, 10, 11])).asnumpy())
assert np.all(b.asnumpy() == Tensor(np.array([1., 2., 3., 4., 5.])).asnumpy())
assert np.all(c.asnumpy() == Tensor(np.array([2. + 6.j, 3.5 + 5.j])).asnumpy())
@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.platform_arm_ascend_training
@pytest.mark.platform_x86_ascend_training
@pytest.mark.env_onecard
def test_np_array_advanced_index_3():
"""
Feature: JIT Fallback
Description: Test numpy with array advanced index in graph mode.
Expectation: No exception.
"""
@ms_function
def np_array_advanced_index_3():
x = np.arange(32).reshape((8, 4))
a = Tensor(x[[4, 2, 1, 7]])
y = np.arange(32).reshape((8, 4))
b = Tensor(y[[-4, -2, -1, -7]])
z = np.arange(32).reshape((8, 4))
c = Tensor(z[np.ix_([1, 5, 7, 2], [0, 3, 1, 2])])
return a, b, c
a, b, c = np_array_advanced_index_3()
print("a:", a)
print("b:", b)
print("c:", c)

View File

@ -15,12 +15,11 @@
""" test graph fallback """
import pytest
import numpy as np
from mindspore import ms_function, context
from mindspore import ms_function, context, Tensor
context.set_context(mode=context.GRAPH_MODE)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_array_1():
"""
Feature: JIT Fallback
@ -30,12 +29,12 @@ def test_np_array_1():
@ms_function
def np_array_1():
a = np.array([1, 2, 3])
return a
return Tensor(a)
res = np_array_1()
assert res == (1, 2, 3)
expect_res = Tensor(np.array([1, 2, 3]))
assert np.all(res.asnumpy() == expect_res.asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_array_2():
"""
Feature: JIT Fallback
@ -45,12 +44,12 @@ def test_np_array_2():
@ms_function
def np_array_2():
a = np.array([[1, 2], [3, 4]])
return a
return Tensor(a)
res = np_array_2()
assert res == ([1, 2], [3, 4])
expect_res = Tensor(np.array([[1, 2], [3, 4]]))
assert np.all(res.asnumpy() == expect_res.asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_array_3():
"""
Feature: JIT Fallback
@ -60,9 +59,10 @@ def test_np_array_3():
@ms_function
def np_array_3():
a = np.array([1, 2, 3, 4, 5], ndmin=2)
return a
return Tensor(a)
res = np_array_3()
assert res == ([1, 2, 3, 4, 5],)
expect_res = Tensor(np.array([[1, 2, 3, 4, 5]]))
assert np.all(res.asnumpy() == expect_res.asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
@ -75,12 +75,11 @@ def test_np_array_4():
@ms_function
def np_array_4():
a = np.array([1, 2, 3], dtype=complex)
return a
return Tensor(a)
res = np_array_4()
assert res == ((1+0j), (2+0j), (3+0j))
assert np.all(res.asnumpy() == Tensor(np.array([1+0j, 2+0j, 3+0j])).asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_dtype_1():
"""
Feature: JIT Fallback
@ -90,12 +89,11 @@ def test_np_dtype_1():
@ms_function
def np_dtype_1():
t = np.dtype(np.int32)
return t
return Tensor(np.array([1, 2, 3], dtype=t))
res = np_dtype_1()
print("res:", res)
assert np.all(res.asnumpy() == Tensor(np.array([1, 2, 3], dtype=np.int32)).asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_dtype_2():
"""
Feature: JIT Fallback
@ -105,9 +103,9 @@ def test_np_dtype_2():
@ms_function
def np_dtype_2():
t = np.dtype('i4')
return t
return Tensor(np.array([1, 2, 3], dtype=t))
res = np_dtype_2()
print("res:", res)
assert np.all(res.asnumpy() == Tensor(np.array([1, 2, 3], dtype=np.int32)).asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
@ -120,7 +118,7 @@ def test_np_dtype_3():
@ms_function
def np_dtype_3():
t = np.dtype([('age', np.int8)])
return t
return Tensor(np.array([1, 2, 3], dtype=t))
res = np_dtype_3()
print("res:", res)
@ -136,12 +134,11 @@ def test_np_dtype_4():
def np_dtype_4():
student = np.dtype([('name', 'S20'), ('age', 'i1'), ('marks', 'f4')])
a = np.array([('abc', 21, 50), ('xyz', 18, 75)], dtype=student)
return a
return Tensor(a)
res = np_dtype_4()
print("res:", res)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_array_ndim():
"""
Feature: JIT Fallback
@ -151,9 +148,9 @@ def test_np_array_ndim():
@ms_function
def np_array_ndim():
a = np.arange(24)
return a.ndim
return Tensor(a.ndim)
res = np_array_ndim()
print("res:", res)
assert res == 1
@pytest.mark.skip(reason='Not support graph fallback feature yet')
@ -167,9 +164,9 @@ def test_np_array_reshape_1():
def np_array_reshape_1():
a = np.array([[1, 2, 3], [4, 5, 6]])
b = a.reshape(3, 2)
return b.ndim
return Tensor(b.ndim)
res = np_array_reshape_1()
print("res:", res)
assert res == 2
@pytest.mark.skip(reason='Not support graph fallback feature yet')
@ -198,9 +195,10 @@ def test_np_array_itemsize():
@ms_function
def np_array_itemsize():
a = np.array([1, 2, 3, 4, 5], dtype=np.int8)
return a.itemsize
return Tensor(a.itemsize)
res = np_array_itemsize()
print("res:", res)
assert res == 1
@pytest.mark.skip(reason='Not support graph fallback feature yet')
@ -218,7 +216,6 @@ def test_np_array_flags():
print("res:", res)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_empty_zeros_ones():
"""
Feature: JIT Fallback
@ -230,12 +227,12 @@ def test_np_empty_zeros_ones():
x = np.empty([3, 2], dtype=np.int)
y = np.zeros(x.shape, dtype=np.int)
z = np.ones(x.shape, dtype=np.int)
return y + z
return Tensor(y + z)
res = np_empty_zeros_ones()
print("res:", res)
except_res = Tensor(np.ones([3, 2], dtype=np.int))
assert np.all(res.asnumpy() == except_res.asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_asarray_list():
"""
Feature: JIT Fallback
@ -246,12 +243,12 @@ def test_np_asarray_list():
def np_asarray_list():
x = [1, 2, 3]
y = np.asarray(x)
return y
return Tensor(y)
res = np_asarray_list()
print("res:", res)
except_res = Tensor(np.asarray([1, 2, 3]))
assert np.all(res.asnumpy() == except_res.asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_asarray_tuple():
"""
Feature: JIT Fallback
@ -262,9 +259,10 @@ def test_np_asarray_tuple():
def np_asarray_tuple():
x = (1, 2, 3)
y = np.asarray(x)
return y
return Tensor(y)
res = np_asarray_tuple()
print("res:", res)
except_res = Tensor(np.asarray((1, 2, 3)))
assert np.all(res.asnumpy() == except_res.asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
@ -278,7 +276,7 @@ def test_np_asarray_tuple_list():
def np_asarray_tuple_list():
x = [(1, 2, 3), (4, 5)]
y = np.asarray(x)
return y
return Tensor(y)
res = np_asarray_tuple_list()
print("res:", res)
@ -299,7 +297,6 @@ def test_np_frombuffer():
print("res:", res)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_fromiter():
"""
Feature: JIT Fallback
@ -311,12 +308,12 @@ def test_np_fromiter():
l = range(5)
it = iter(l)
x = np.fromiter(it, dtype=float)
return x
return Tensor(x)
res = np_fromiter()
print("res:", res)
except_res = Tensor(np.asarray([0., 1., 2., 3., 4.]))
assert np.all(res.asnumpy() == except_res.asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_arange():
"""
Feature: JIT Fallback
@ -327,38 +324,12 @@ def test_np_arange():
def np_arange():
x = np.arange(5, dtype=float)
y = np.arange(10, 20, 2)
return x, y
res1, res2 = np_arange()
print("res1:", res1)
print("res2:", res2)
return Tensor(x + y)
res = np_arange()
except_res = Tensor(np.asarray([10., 13., 16., 19., 22.]))
assert np.all(res.asnumpy() == except_res.asnumpy())
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_linspace():
"""
Feature: JIT Fallback
Description: Test numpy with linspace in graph mode.
Expectation: No exception.
"""
@ms_function
def np_linspace():
a = np.linspace(1, 10, 10)
b = np.linspace(1, 1, 10)
c = np.linspace(10, 20, 5, endpoint=False)
d = np.linspace(10, 20, 5, endpoint=True)
e = np.linspace(1, 10, 10, retstep=True)
f = np.linspace(1, 10, 10).reshape([10, 1])
return a, b, c, d, e, f
a, b, c, d, e, f = np_linspace()
print("a:", a)
print("b:", b)
print("c:", c)
print("d:", d)
print("e:", e)
print("f:", f)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_logspace():
"""
Feature: JIT Fallback
@ -367,126 +338,8 @@ def test_np_logspace():
"""
@ms_function
def np_logspace():
a = np.logspace(1.0, 2.0, num=10)
b = np.logspace(0, 9, 10, base=2)
return a, b
a, b = np_logspace()
print("a:", a)
print("b:", b)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_arange_slice_1():
"""
Feature: JIT Fallback
Description: Test numpy with arange slice in graph mode.
Expectation: No exception.
"""
@ms_function
def np_arange_slice_1():
x = np.arange(10)
index = slice(2, 7, 2)
a = x[index]
b = x[2:7:2]
c = x[5]
d = x[2:]
e = x[2:5]
return a, b, c, d, e
a, b, c, d, e = np_arange_slice_1()
print("a:", a)
print("b:", b)
print("c:", c)
print("d:", d)
print("e:", e)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_arange_slice_2():
"""
Feature: JIT Fallback
Description: Test numpy with arange slice in graph mode.
Expectation: No exception.
"""
@ms_function
def np_arange_slice_2():
x = np.array([[1, 2, 3], [3, 4, 5], [4, 5, 6]])
a = x[1:]
b = x[..., 1]
c = x[1, ...]
d = x[..., 1:]
return a, b, c, d
a, b, c, d = np_arange_slice_2()
print("a:", a)
print("b:", b)
print("c:", c)
print("d:", d)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_array_advanced_index_1():
"""
Feature: JIT Fallback
Description: Test numpy with array advanced index in graph mode.
Expectation: No exception.
"""
@ms_function
def np_array_advanced_index_1():
x = np.array([[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11]])
a = x[[0, 1, 2], [0, 1, 0]]
rows = np.array([[0, 0], [3, 3]])
cols = np.array([[0, 2], [0, 2]])
b = x[rows, cols]
c = x[1:3, 1:3]
d = x[1:3, [1, 2]]
e = x[..., 1:]
return a, b, c, d, e
a, b, c, d, e = np_array_advanced_index_1()
print("a:", a)
print("b:", b)
print("c:", c)
print("d:", d)
print("e:", e)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_array_advanced_index_2():
"""
Feature: JIT Fallback
Description: Test numpy with array advanced index in graph mode.
Expectation: No exception.
"""
@ms_function
def np_array_advanced_index_2():
x = np.array([[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11]])
y = np.array([np.nan, 1, 2, np.nan, 3, 4, 5])
z = np.array([1, 2 + 6j, 5, 3.5 + 5j])
a = x[x > 5]
b = y[~np.isnan(y)]
c = z[np.iscomplex(z)]
return a, b, c
a, b, c = np_array_advanced_index_2()
print("a:", a)
print("b:", b)
print("c:", c)
@pytest.mark.skip(reason='Not support graph fallback feature yet')
def test_np_array_advanced_index_3():
"""
Feature: JIT Fallback
Description: Test numpy with array advanced index in graph mode.
Expectation: No exception.
"""
@ms_function
def np_array_advanced_index_3():
x = np.arange(32).reshape((8, 4))
a = x[[4, 2, 1, 7]]
y = np.arange(32).reshape((8, 4))
b = y[[-4, -2, -1, -7]]
z = np.arange(32).reshape((8, 4))
c = z[np.ix_([1, 5, 7, 2], [0, 3, 1, 2])]
return a, b, c
a, b, c = np_array_advanced_index_3()
print("a:", a)
print("b:", b)
print("c:", c)
a = np.logspace(0, 9, 10, base=2)
return Tensor(a)
res = np_logspace()
except_res = Tensor(np.array([1., 2., 4., 8., 16., 32., 64., 128., 256., 512.]))
assert np.all(res.asnumpy() == except_res.asnumpy())