forked from mindspore-Ecosystem/mindspore
73 lines
2.5 KiB
Python
73 lines
2.5 KiB
Python
# 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.
|
|
# ============================================================================
|
|
|
|
import numpy as np
|
|
import pytest
|
|
import mindspore.context as context
|
|
from mindspore import Tensor
|
|
from mindspore.nn import Cell
|
|
import mindspore.ops.operations as P
|
|
import mindspore.common.dtype as mstype
|
|
|
|
|
|
class FusionNet(Cell):
|
|
def __init__(self):
|
|
super(FusionNet, self).__init__()
|
|
self.onehot = P.OneHot()
|
|
self.on_value = Tensor(1.0, mstype.float32)
|
|
self.off_value = Tensor(0.0, mstype.float32)
|
|
self.add = P.Add()
|
|
self.reshape = P.Reshape()
|
|
self.mul = P.Mul()
|
|
|
|
def construct(self, x, y, indices, depth):
|
|
res_1 = self.reshape(indices, (4096,))
|
|
res_2 = self.onehot(res_1, depth, self.on_value, self.off_value)
|
|
res_3 = self.mul(res_2, x)
|
|
res_4 = self.add(res_3, y)
|
|
|
|
return res_4
|
|
|
|
|
|
def fusion_net_get_output(x, y, indices, depth, enable_graph_kernel=False):
|
|
context.set_context(enable_graph_kernel=enable_graph_kernel)
|
|
net = FusionNet()
|
|
|
|
output = net(x, y, indices, depth)
|
|
return output
|
|
|
|
|
|
def fusion_net_compare_result():
|
|
depth = 512
|
|
indices = Tensor(np.random.randint(depth, size=[4, 1024]).astype(np.int32))
|
|
x = Tensor(np.random.normal(0, 1, [4096, 512]).astype(np.float32))
|
|
y = Tensor(np.random.normal(0, 1, [4096, 1]).astype(np.float32))
|
|
expect = fusion_net_get_output(x, y, indices, depth, False)
|
|
output = fusion_net_get_output(x, y, indices, depth, True)
|
|
assert np.allclose(expect.asnumpy(), output.asnumpy(), 1.e-4, 1.e-7)
|
|
|
|
|
|
@pytest.mark.level0
|
|
@pytest.mark.platform_x86_gpu_training
|
|
@pytest.mark.env_onecard
|
|
def test_gpu_graph_mode():
|
|
"""
|
|
Feature: graph kernel testcase for onehot
|
|
Description: random input when using graph_kernel in graph mode
|
|
Expectation: get the same result when using and not using graph kernel
|
|
"""
|
|
context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
|
|
fusion_net_compare_result()
|