forked from mindspore-Ecosystem/mindspore
130 lines
4.5 KiB
Python
130 lines
4.5 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 mindspore.context as context
|
|
import mindspore.nn as nn
|
|
from mindspore import Tensor
|
|
from mindspore.ops.operations.inner_ops import ScaleGrad
|
|
from mindspore.common import dtype
|
|
|
|
context.set_context(device_target="Ascend")
|
|
|
|
|
|
class Net(nn.Cell):
|
|
def __init__(self, scale):
|
|
super(Net, self).__init__()
|
|
self.scale_grad = ScaleGrad()
|
|
self.scale = scale
|
|
|
|
def construct(self, origin_grads):
|
|
return self.scale_grad(origin_grads, self.scale)
|
|
|
|
|
|
def test_scale_grad_grad_float32_scale_float32():
|
|
"""
|
|
Feature: Scale Grad fusion operation
|
|
Description: test the rightness of ScaleGrad kernel, gradient's dtype is float32, scale's dtype is float32
|
|
Expectation: the output is wrong
|
|
"""
|
|
scale = Tensor(1024.0, dtype.float32)
|
|
gradients = []
|
|
for _ in range(3):
|
|
gradients.append(Tensor(np.ones([3, 3]).astype(np.float32)))
|
|
gradients_input = tuple(gradients)
|
|
scale_grad = Net(scale)
|
|
scale_grad(gradients_input)
|
|
|
|
|
|
def test_scale_grad_grad_float32_scale_float16():
|
|
"""
|
|
Feature: Scale Grad fusion operation
|
|
Description: test the rightness of ScaleGrad kernel, gradient's dtype is float32, scale's dtype is float16
|
|
Expectation: the output is wrong
|
|
"""
|
|
scale = Tensor(1024.0, dtype.float32)
|
|
gradients = []
|
|
for _ in range(3):
|
|
gradients.append(Tensor(np.ones([3, 3]).astype(np.float32)))
|
|
gradients_input = tuple(gradients)
|
|
scale_grad = Net(scale)
|
|
scale_grad(gradients_input)
|
|
|
|
|
|
def test_scale_grad_grad_float16_scale_float32():
|
|
"""
|
|
Feature: Scale Grad fusion operation
|
|
Description: test the rightness of ScaleGrad kernel, gradient's dtype is float16, scale's dtype is float32
|
|
Expectation: the output is wrong
|
|
"""
|
|
scale = Tensor(1024.0, dtype.float32)
|
|
gradients = []
|
|
for _ in range(3):
|
|
gradients.append(Tensor(np.ones([3, 3]).astype(np.float16)))
|
|
gradients_input = tuple(gradients)
|
|
scale_grad = Net(scale)
|
|
scale_grad(gradients_input)
|
|
|
|
|
|
def test_scale_grad_grad_float16_scale_float16():
|
|
"""
|
|
Feature: Scale Grad fusion operation
|
|
Description: test the rightness of ScaleGrad kernel, gradient's dtype is float16, scale's dtype is float16
|
|
Expectation: the output is wrong
|
|
"""
|
|
scale = Tensor(1024.0, dtype.float16)
|
|
gradients = []
|
|
for _ in range(3):
|
|
gradients.append(Tensor(np.ones([3, 3]).astype(np.float16)))
|
|
gradients_input = tuple(gradients)
|
|
scale_grad = Net(scale)
|
|
scale_grad(gradients_input)
|
|
|
|
|
|
def test_scale_grad_grad_mixed_scale_float32():
|
|
"""
|
|
Feature: Scale Grad fusion operation
|
|
Description: test the rightness of ScaleGrad kernel, gradient's dtype is mixed, scale's dtype is float32
|
|
Expectation: the output is wrong
|
|
"""
|
|
scale = Tensor(1024.0, dtype.float32)
|
|
gradients = []
|
|
for i in range(3):
|
|
if (i % 2) == 0:
|
|
gradients.append(Tensor(np.ones([3, 3]).astype(np.float32)))
|
|
else:
|
|
gradients.append(Tensor(np.ones([3, 3]).astype(np.float16)))
|
|
gradients_input = tuple(gradients)
|
|
scale_grad = Net(scale)
|
|
scale_grad(gradients_input)
|
|
|
|
|
|
def test_scale_grad_grad_mixed_scale_float16():
|
|
"""
|
|
Feature: Scale Grad fusion operation
|
|
Description: test the rightness of ScaleGrad kernel, gradient's dtype is mixed, scale's dtype is float16
|
|
Expectation: the output is wrong
|
|
"""
|
|
scale = Tensor(1024.0, dtype.float16)
|
|
gradients = []
|
|
for i in range(3):
|
|
if (i % 2) == 0:
|
|
gradients.append(Tensor(np.ones([3, 3]).astype(np.float32)))
|
|
else:
|
|
gradients.append(Tensor(np.ones([3, 3]).astype(np.float16)))
|
|
gradients_input = tuple(gradients)
|
|
scale_grad = Net(scale)
|
|
scale_grad(gradients_input)
|