1# Copyright 2020 Huawei Technologies Co., Ltd 2# 3# Licensed under the Apache License, Version 2.0 (the "License"); 4# you may not use this file except in compliance with the License. 5# You may obtain a copy of the License at 6# 7# http://www.apache.org/licenses/LICENSE-2.0 8# 9# Unless required by applicable law or agreed to in writing, software 10# distributed under the License is distributed on an "AS IS" BASIS, 11# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 12# See the License for the specific language governing permissions and 13# limitations under the License. 14# ============================================================================ 15"""test lenet""" 16import numpy as np 17 18import mindspore.context as context 19import mindspore.nn as nn 20from mindspore import Tensor 21from mindspore.common.api import _cell_graph_executor 22from mindspore.ops import operations as P 23from ....train_step_wrap import train_step_with_loss_warp, train_step_with_sens 24 25context.set_context(mode=context.GRAPH_MODE) 26 27 28class LeNet5(nn.Cell): 29 """LeNet5 definition""" 30 31 def __init__(self): 32 super(LeNet5, self).__init__() 33 self.conv1 = nn.Conv2d(1, 6, 5, pad_mode='valid') 34 self.conv2 = nn.Conv2d(6, 16, 5, pad_mode='valid') 35 self.fc1 = nn.Dense(16 * 5 * 5, 120) 36 self.fc2 = nn.Dense(120, 84) 37 self.fc3 = nn.Dense(84, 10) 38 self.relu = nn.ReLU() 39 self.max_pool2d = nn.MaxPool2d(kernel_size=2, stride=2) 40 self.flatten = P.Flatten() 41 42 def construct(self, x): 43 x = self.max_pool2d(self.relu(self.conv1(x))) 44 x = self.max_pool2d(self.relu(self.conv2(x))) 45 x = self.flatten(x) 46 x = self.relu(self.fc1(x)) 47 x = self.relu(self.fc2(x)) 48 x = self.fc3(x) 49 return x 50 51 52def test_lenet5_train_step(): 53 predict = Tensor(np.ones([1, 1, 32, 32]).astype(np.float32) * 0.01) 54 label = Tensor(np.zeros([1, 10]).astype(np.float32)) 55 net = train_step_with_loss_warp(LeNet5()) 56 _cell_graph_executor.compile(net, predict, label) 57 58 59def test_lenet5_train_sens(): 60 predict = Tensor(np.ones([1, 1, 32, 32]).astype(np.float32) * 0.01) 61 sens = Tensor(np.ones([1, 10]).astype(np.float32)) 62 net = train_step_with_sens(LeNet5(), sens) 63 _cell_graph_executor.compile(net, predict) 64 65 66def test_lenet5_train_step_training(): 67 predict = Tensor(np.ones([1, 1, 32, 32]).astype(np.float32) * 0.01) 68 label = Tensor(np.zeros([1, 10]).astype(np.float32)) 69 net = train_step_with_loss_warp(LeNet5()) 70 net.set_train() 71 _cell_graph_executor.compile(net, predict, label) 72