mindspore/tests/st/networks/test_gpu_lenet.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 os
import pytest
import numpy as np
from mindspore import Tensor
import mindspore.context as context
from mindspore.ops import operations as P
import mindspore.nn as nn
from mindspore.nn import Dense, TrainOneStepCell, WithLossCell
from mindspore.nn.optim import Momentum
from mindspore.nn.metrics import Accuracy
from mindspore.train import Model
from mindspore.common import dtype as mstype
from mindspore.common.initializer import initializer
from mindspore.model_zoo.lenet import LeNet5
from mindspore.train.callback import LossMonitor

import mindspore.dataset as ds
import mindspore.dataset.transforms.vision.c_transforms as CV
import mindspore.dataset.transforms.c_transforms as C
from mindspore.dataset.transforms.vision import Inter

context.set_context(mode=context.GRAPH_MODE, device_target="GPU")


class LeNet(nn.Cell):
    def __init__(self):
        super(LeNet, self).__init__()
        self.relu = P.ReLU()
        self.batch_size = 1
        weight1 = Tensor(np.ones([6, 3, 5, 5]).astype(np.float32) * 0.01)
        weight2 = Tensor(np.ones([16, 6, 5, 5]).astype(np.float32) * 0.01)
        self.conv1 = nn.Conv2d(3, 6, (5, 5), weight_init=weight1, stride=1, padding=0, pad_mode='valid')
        self.conv2 = nn.Conv2d(6, 16, (5, 5), weight_init=weight2, pad_mode='valid', stride=1, padding=0)
        self.pool = nn.MaxPool2d(kernel_size=2, stride=2, pad_mode="valid")

        self.reshape = P.Reshape()
        self.reshape1 = P.Reshape()

        self.fc1 = Dense(400, 120)
        self.fc2 = Dense(120, 84)
        self.fc3 = Dense(84, 10)

    def construct(self, input_x):
        output = self.conv1(input_x)
        output = self.relu(output)
        output = self.pool(output)
        output = self.conv2(output)
        output = self.relu(output)
        output = self.pool(output)
        output = self.reshape(output, (self.batch_size, -1))
        output = self.fc1(output)
        output = self.fc2(output)
        output = self.fc3(output)
        return output


def multisteplr(total_steps, gap, base_lr=0.9, gamma=0.1, dtype=mstype.float32):
    lr = []
    for step in range(total_steps):
        lr_ = base_lr * gamma ** (step // gap)
        lr.append(lr_)
    return Tensor(np.array(lr), dtype)


@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_train_lenet():
    epoch = 100
    net = LeNet()
    momentum = initializer(Tensor(np.array([0.9]).astype(np.float32)), [1])
    learning_rate = multisteplr(epoch, 30)

    optimizer = Momentum(filter(lambda x: x.requires_grad, net.get_parameters()), learning_rate, momentum)
    criterion = nn.SoftmaxCrossEntropyWithLogits(is_grad=False, sparse=True)
    net_with_criterion = WithLossCell(net, criterion)
    train_network = TrainOneStepCell(net_with_criterion, optimizer)  # optimizer
    train_network.set_train()
    losses = []
    for i in range(epoch):
        data = Tensor(np.ones([net.batch_size, 3, 32, 32]).astype(np.float32) * 0.01)
        label = Tensor(np.ones([net.batch_size]).astype(np.int32))
        loss = train_network(data, label)
        losses.append(loss)
    print(losses)


def create_dataset(data_path, batch_size=32, repeat_size=1,
                   num_parallel_workers=1):
    """
    create dataset for train or test
    """
    # define dataset
    mnist_ds = ds.MnistDataset(data_path)

    resize_height, resize_width = 32, 32
    rescale = 1.0 / 255.0
    shift = 0.0
    rescale_nml = 1 / 0.3081
    shift_nml = -1 * 0.1307 / 0.3081

    # define map operations
    resize_op = CV.Resize((resize_height, resize_width), interpolation=Inter.LINEAR)  # Bilinear mode
    rescale_nml_op = CV.Rescale(rescale_nml, shift_nml)
    rescale_op = CV.Rescale(rescale, shift)
    hwc2chw_op = CV.HWC2CHW()
    type_cast_op = C.TypeCast(mstype.int32)

    # apply map operations on images
    mnist_ds = mnist_ds.map(input_columns="label", operations=type_cast_op, num_parallel_workers=num_parallel_workers)
    mnist_ds = mnist_ds.map(input_columns="image", operations=resize_op, num_parallel_workers=num_parallel_workers)
    mnist_ds = mnist_ds.map(input_columns="image", operations=rescale_op, num_parallel_workers=num_parallel_workers)
    mnist_ds = mnist_ds.map(input_columns="image", operations=rescale_nml_op, num_parallel_workers=num_parallel_workers)
    mnist_ds = mnist_ds.map(input_columns="image", operations=hwc2chw_op, num_parallel_workers=num_parallel_workers)

    # apply DatasetOps
    buffer_size = 10000
    mnist_ds = mnist_ds.shuffle(buffer_size=buffer_size)  # 10000 as in LeNet train script
    mnist_ds = mnist_ds.batch(batch_size, drop_remainder=True)
    mnist_ds = mnist_ds.repeat(repeat_size)

    return mnist_ds


@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_train_and_eval_lenet():
    context.set_context(mode=context.GRAPH_MODE, device_target="GPU", enable_mem_reuse=False)
    network = LeNet5(10)
    net_loss = nn.SoftmaxCrossEntropyWithLogits(is_grad=False, sparse=True, reduction="mean")
    net_opt = nn.Momentum(network.trainable_params(), 0.01, 0.9)
    model = Model(network, net_loss, net_opt, metrics={"Accuracy": Accuracy()})

    print("============== Starting Training ==============")
    ds_train = create_dataset(os.path.join('/home/workspace/mindspore_dataset/mnist', "train"), 32, 1)
    model.train(1, ds_train, callbacks=[LossMonitor()], dataset_sink_mode=True)

    print("============== Starting Testing ==============")
    ds_eval = create_dataset(os.path.join('/home/workspace/mindspore_dataset/mnist', "test"), 32, 1)
    acc = model.eval(ds_eval, dataset_sink_mode=True)
    print("============== Accuracy:{} ==============".format(acc))