test_alltoall.py

# 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
from mindspore import context
from mindspore.common.api import _executor
from mindspore.common.parameter import Parameter
from mindspore.nn.loss import SoftmaxCrossEntropyWithLogits
from mindspore.nn.optim.momentum import Momentum
from mindspore.ops import operations as P
from mindspore.parallel._utils import _reset_op_id
from mindspore.train import Model
from mindspore.context import ParallelMode
from tests.dataset_mock import MindData


class Dataset(MindData):
    def __init__(self, predict, label, length=3):
        super(Dataset, self).__init__(size=length)
        self.predict = predict
        self.label = label
        self.index = 0
        self.length = length

    def __iter__(self):
        return self

    def __next__(self):
        if self.index >= self.length:
            raise StopIteration
        self.index += 1
        return self.predict, self.label

    def reset(self):
        self.index = 0


class AllToAllNet(nn.Cell):
    def __init__(self, strategy1):
        super(AllToAllNet, self).__init__()
        self.matmul = P.MatMul().set_strategy(((1, 1), (1, 8)))
        self.matmul_weight = Parameter(Tensor(np.ones([128, 256]), dtype=ms.float32), name="weight")
        self.transpose1 = P.Transpose().set_strategy(strategy1)

    def construct(self, x):
        x = self.matmul(x, self.matmul_weight)
        x = self.transpose1(x, (1, 0))
        return x


def all_to_all_net(strategy1):
    return AllToAllNet(strategy1=strategy1)


def all_to_all_common(strategy1):
    learning_rate = 0.1
    momentum = 0.9
    epoch_size = 2

    context.reset_auto_parallel_context()
    context.set_auto_parallel_context(parallel_mode=ParallelMode.SEMI_AUTO_PARALLEL, device_num=8)
    predict = Tensor(np.ones([32, 128]), dtype=ms.float32)
    label = Tensor(np.ones([32]), dtype=ms.int32)
    dataset = Dataset(predict, label, 2)
    net = all_to_all_net(strategy1)

    loss = SoftmaxCrossEntropyWithLogits(sparse=True)
    loss.softmax_cross_entropy.set_strategy(((8, 1), (8, 1)))
    loss.one_hot.set_strategy(((8, 1), (), ()))
    opt = Momentum(net.trainable_params(), learning_rate, momentum)
    model = Model(net, loss, opt)

    model.train(epoch_size, dataset, dataset_sink_mode=False)
    strategys = _executor._get_strategy(model._train_network)
    return strategys


def test_all_to_all():
    strategy1 = ((8, 1),)
    context.set_context(mode=context.GRAPH_MODE, save_graphs=False)
    _reset_op_id()
    strategys = all_to_all_common(strategy1)
    print(strategys)
    expect_dict = {'Default/network-_VirtualDatasetCell/_backbone-WithLossCell/_loss_fn-SoftmaxCrossEntropyWithLogits'
                   '/SoftmaxCrossEntropyWithLogits-op3': [[8, 1], [8, 1]],
                   'Default/network-_VirtualDatasetCell/_backbone-WithLossCell/_loss_fn-SoftmaxCrossEntropyWithLogits/'
                   'OneHot-op4': [[8, 1], [], []],
                   'Default/network-_VirtualDatasetCell/_backbone-WithLossCell/_backbone-AllToAllNet/Transpose-op1': [
                       [8, 1]],
                   'Default/network-_VirtualDatasetCell/_backbone-WithLossCell/_backbone-AllToAllNet/MatMul-op0': [
                       [1, 1], [1, 8]]}
    assert strategys == expect_dict
    context.set_context(save_graphs=False)


if __name__ == '__main__':
    test_all_to_all()