mindspore/tests/st/ops/gpu/test_xdivy_op.py

188 lines
6.4 KiB
Python

# Copyright 2020-2022 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.
# ============================================================================
import numpy as np
import pytest
import mindspore.context as context
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.ops import operations as P
import mindspore as ms
TF_INSTALL_FLG = 1
try:
import tensorflow as tf
except ImportError:
TF_INSTALL_FLG = 0
class NetXDivy(nn.Cell):
def __init__(self):
super(NetXDivy, self).__init__()
self.xdivy = P.Xdivy()
def construct(self, x, y):
return self.xdivy(x, y)
def xdivy(nptype):
x0_np = np.random.randint(1, 5, (2, 3, 4, 4)).astype(nptype)
y0_np = np.random.randint(1, 5, (2, 3, 4, 4)).astype(nptype)
x1_np = np.random.randint(1, 5, (2, 3, 4, 4)).astype(nptype)
y1_np = np.random.randint(1, 5, (2, 1, 4, 4)).astype(nptype)
x2_np = np.random.randint(1, 5, (2, 1, 1, 4)).astype(nptype)
y2_np = np.random.randint(1, 5, (2, 3, 4, 4)).astype(nptype)
x3_np = np.random.randint(1, 5, 1).astype(nptype)
y3_np = np.random.randint(1, 5, 1).astype(nptype)
x4_np = np.array(78).astype(nptype)
y4_np = np.array(37.5).astype(nptype)
x0 = Tensor(x0_np)
y0 = Tensor(y0_np)
x1 = Tensor(x1_np)
y1 = Tensor(y1_np)
x2 = Tensor(x2_np)
y2 = Tensor(y2_np)
x3 = Tensor(x3_np)
y3 = Tensor(y3_np)
x4 = Tensor(x4_np)
y4 = Tensor(y4_np)
context.set_context(mode=context.GRAPH_MODE, device_target='GPU')
div_net = NetXDivy()
output0 = div_net(x0, y0)
expect0 = np.divide(x0_np, y0_np)
diff0 = output0.asnumpy() - expect0
error0 = np.ones(shape=expect0.shape) * 1.0e-5
assert np.all(diff0 < error0)
assert output0.shape == expect0.shape
output1 = div_net(x1, y1)
expect1 = np.divide(x1_np, y1_np)
diff1 = output1.asnumpy() - expect1
error1 = np.ones(shape=expect1.shape) * 1.0e-5
assert np.all(diff1 < error1)
assert output1.shape == expect1.shape
output2 = div_net(x2, y2)
expect2 = np.divide(x2_np, y2_np)
diff2 = output2.asnumpy() - expect2
error2 = np.ones(shape=expect2.shape) * 1.0e-5
assert np.all(diff2 < error2)
assert output2.shape == expect2.shape
context.set_context(mode=context.PYNATIVE_MODE, device_target='GPU')
output3 = div_net(x3, y3)
expect3 = np.divide(x3_np, y3_np)
diff3 = output3.asnumpy() - expect3
error3 = np.ones(shape=expect3.shape) * 1.0e-5
assert np.all(diff3 < error3)
assert output3.shape == expect3.shape
output4 = div_net(x4, y4)
expect4 = np.divide(x4_np, y4_np)
diff4 = output4.asnumpy() - expect4
error4 = np.ones(shape=expect4.shape) * 1.0e-5
assert np.all(diff4 < error4)
assert output4.shape == expect4.shape
def xdivy_sf_check(mstype, tftype):
# test divided zero
with tf.device('/cpu:0'):
tx = tf.constant([-4.0, 0.0, 1.0, 0.0], dtype=tftype)
ty = tf.constant([3.0, 2.0, 0.0, 0.0], dtype=tftype)
tz = tf.math.xdivy(tx, ty)
x = ms.Tensor(np.array([-4.0, 0.0, 1.0, 0.0]), dtype=mstype)
y = ms.Tensor(np.array([3.0, 2.0, 0.0, 0.0]), dtype=mstype)
z = ms.ops.xdivy(x, y)
assert tz.numpy().all() == z.asnumpy().all()
# test broadcast
with tf.device('/cpu:0'):
tx = tf.constant([-4.0, 5.0, 0.0], dtype=tftype)
ty = tf.constant([[3.0], [2.0]], dtype=tftype)
tz = tf.math.xdivy(tx, ty)
x = ms.Tensor(np.array([-4.0, 5.0, 0.0]), dtype=mstype)
y = ms.Tensor(np.array([[3.0], [2.0]]), dtype=mstype)
z = ms.ops.xdivy(x, y)
assert tz.numpy().all() == z.asnumpy().all()
# test broadcast
with tf.device('/cpu:0'):
tx = tf.constant([-4.0], dtype=tftype)
ty = tf.constant([[3.0, 1.0, 1.0], [2.0, 3.0, 5.0]], dtype=tftype)
tz = tf.math.xdivy(tx, ty)
x = ms.Tensor(np.array([-4.0]), dtype=mstype)
y = ms.Tensor(np.array([[3.0, 1.0, 1.0], [2.0, 3.0, 5.0]]), dtype=mstype)
z = ms.ops.xdivy(x, y)
assert tz.numpy().all() == z.asnumpy().all()
@pytest.mark.level1
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_div_float64():
"""
Feature: test xdivy primitive use float64
Description: compare result with numpy&& tensorflow
Expectation: calculate result same to numpy&&tensorflow
"""
xdivy(np.float64)
if TF_INSTALL_FLG == 0:
return
context.set_context(mode=context.PYNATIVE_MODE, device_target='GPU')
xdivy_sf_check(ms.float64, tf.dtypes.float64)
context.set_context(mode=context.GRAPH_MODE, device_target='GPU')
xdivy_sf_check(ms.float64, tf.dtypes.float64)
@pytest.mark.level1
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_div_float32():
"""
Feature: test xdivy primitive use float32
Description: compare result with numpy&& tensorflow
Expectation: calculate result same to numpy&&tensorflow
"""
xdivy(np.float32)
if TF_INSTALL_FLG == 0:
return
context.set_context(mode=context.PYNATIVE_MODE, device_target='GPU')
xdivy_sf_check(ms.float32, tf.dtypes.float32)
context.set_context(mode=context.GRAPH_MODE, device_target='GPU')
xdivy_sf_check(ms.float32, tf.dtypes.float32)
@pytest.mark.level1
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_div_float16():
"""
Feature: test xdivy primitive use float16
Description: compare result with numpy&& tensorflow
Expectation: calculate result same to numpy&&tensorflow
"""
xdivy(np.float16)
if TF_INSTALL_FLG == 0:
return
context.set_context(mode=context.PYNATIVE_MODE, device_target='GPU')
xdivy_sf_check(ms.float16, tf.dtypes.float16)
context.set_context(mode=context.GRAPH_MODE, device_target='GPU')
xdivy_sf_check(ms.float16, tf.dtypes.float16)