You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
mindspore/tests/ut/python/parallel/test_auto_parallel_rhombus.py

142 lines
4.7 KiB

# Copyright 2019 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 as ms
import mindspore.nn as nn
from mindspore import Tensor, Parameter
from mindspore import context
from mindspore.common.api import _executor
from mindspore.ops import composite as C
from mindspore.ops import operations as P
from tests.ut.python.ops.test_math_ops import VirtualLoss
grad_all = C.GradOperation(get_all=True)
class NetWithLoss(nn.Cell):
def __init__(self, network):
super(NetWithLoss, self).__init__()
self.loss = VirtualLoss()
self.network = network
def construct(self, x, y, b):
predict = self.network(x, y, b)
return self.loss(predict)
class GradWrap(nn.Cell):
def __init__(self, network):
super(GradWrap, self).__init__()
self.network = network
def construct(self, x, y, b):
return grad_all(self.network)(x, y, b)
def compile_net(net, x, y, b):
net.set_auto_parallel()
net.set_train()
_executor.compile(net, x, y, b)
def test_rhombus1():
class Net(nn.Cell):
def __init__(self):
super().__init__()
self.matmul = P.MatMul()
self.tadd1 = P.Add()
self.tadd2 = P.Add()
self.weight = Parameter(Tensor(np.ones([128, 128]).astype(np.float32) * 0.01), "w", requires_grad=True)
def construct(self, x, y, z):
mm_out = self.matmul(x, self.weight)
ta1_out = self.tadd1(y, z)
out = self.tadd2(ta1_out, mm_out)
return out
size = 16
context.set_auto_parallel_context(device_num=size, global_rank=0)
x = Tensor(np.ones([128, 128]), dtype=ms.float32)
y = Tensor(np.ones([128, 128]), dtype=ms.float32)
b = Tensor(np.ones([128, 128]), dtype=ms.float32)
net = GradWrap(NetWithLoss(Net()))
context.set_auto_parallel_context(parallel_mode="auto_parallel")
compile_net(net, x, y, b)
def test_rhombus2():
class Net(nn.Cell):
def __init__(self):
super().__init__()
self.matmul1 = P.MatMul()
self.matmul2 = P.MatMul()
self.tadd1 = P.Add()
self.tadd2 = P.Add()
self.tadd3 = P.Add()
self.weight1 = Parameter(Tensor(np.ones([128, 128]).astype(np.float32) * 0.01), "w", requires_grad=True)
self.weight2 = Parameter(Tensor(np.ones([128, 128]).astype(np.float32) * 0.01), "w", requires_grad=True)
def construct(self, x, y, z):
mm1_out = self.matmul1(x, self.weight1)
ta1_out = self.tadd1(y, z)
ta2_out = self.tadd2(mm1_out, ta1_out)
mm2_out = self.matmul2(ta1_out, self.weight2)
ta3_out = self.tadd3(ta2_out, mm2_out)
return ta3_out
size = 16
context.set_auto_parallel_context(device_num=size, global_rank=0)
x = Tensor(np.ones([128, 128]), dtype=ms.float32)
y = Tensor(np.ones([128, 128]), dtype=ms.float32)
b = Tensor(np.ones([128, 128]), dtype=ms.float32)
net = GradWrap(NetWithLoss(Net()))
context.set_auto_parallel_context(parallel_mode="auto_parallel")
compile_net(net, x, y, b)
def test_rhombus3():
class Net(nn.Cell):
def __init__(self):
super().__init__()
self.matmul1 = P.MatMul()
self.tadd1 = P.Add()
self.tadd2 = P.Add()
self.tadd3 = P.Add()
self.tadd4 = P.Add()
self.weight1 = Parameter(Tensor(np.ones([128, 128]).astype(np.float32) * 0.01), "w", requires_grad=True)
self.t = Tensor(np.ones([128, 128]).astype(np.float32) * 0.01)
def construct(self, x, y, z):
mm1_out = self.matmul1(x, self.weight1)
ta1_out = self.tadd1(y, z)
ta2_out = self.tadd2(mm1_out, ta1_out)
ta3_out = self.tadd3(ta1_out, self.t)
ta4_out = self.tadd4(ta2_out, ta3_out)
return ta4_out
size = 16
context.set_auto_parallel_context(device_num=size, global_rank=0)
x = Tensor(np.ones([128, 128]), dtype=ms.float32)
y = Tensor(np.ones([128, 128]), dtype=ms.float32)
z = Tensor(np.ones([128, 128]), dtype=ms.float32)
net = GradWrap(NetWithLoss(Net()))
context.set_auto_parallel_context(parallel_mode="auto_parallel")
compile_net(net, x, y, z)