pytorchtool/example.py at main · FENP/pytorchtool · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
import os
import sys
sys.path.append('..')
import time
import torch
import pytorchtool
import numpy as np

from classes import class_names

from PIL import Image
from torchvision import models, transforms

os.environ["CUDA_VISIBLE_DEVICES"] = "0"

def process_img(path_img):
    # hard code
    norm_mean = [0.485, 0.456, 0.406]
    norm_std = [0.229, 0.224, 0.225]
    inference_transform = transforms.Compose([
        transforms.Resize(256),
        transforms.CenterCrop((224, 224)),
        transforms.ToTensor(),
        transforms.Normalize(norm_mean, norm_std),
    ])

    # tensor
    img_tensor = inference_transform(Image.open(path_img).convert('RGB'))
    img_tensor.unsqueeze_(0)        # chw --> bchw

    return img_tensor

class model:
    def __init__(self, model_name, use_gpu=False):
        self.model_name = model_name
        self.x = process_img('./pandas.jpg')
        self.use_gpu = use_gpu

        if self.model_name in 'inception':
            self.model_name = 'inception'
            self.path = "./model_weight/inception_v3/inception_v3_google-1a9a5a14.pth"

            model = models.Inception3(aux_logits=False, transform_input=False,
                                    init_weights=False)
            model.eval()
            self.model = model
            self.depth = 2
        elif self.model_name in 'alexnet':
            self.model_name = 'alexnet'
            self.path = "./model_weight/alexnet/alexnet-owt-4df8aa71.pth"

            model = models.alexnet(False)
            model.eval()
            self.model = model
            self.depth = -1
        elif self.model_name in 'resnet':
            self.model_name = 'resnet'
            self.path = './model_weight/resnet/resnet18-f37072fd.pth'
            model = models.resnet18(False)
            model.eval()
            self.model = model
            self.depth = 2
        else:
            print("Wrong model name")

        if self.use_gpu:
            self.model = self.model.to(0)
            # self.x = self.x.cuda()
            self.x = self.x.to(0)

    def load_weight(self):
        state_dict_read = torch.load(self.path)

        self.model.load_state_dict(state_dict_read, strict=False)

    def get_model(self):
        return self.model

    def get_input(self):
        return self.x

    def save_layers(self, depth=-1):
        pytorchtool.save_model(self.model, depth=depth)

    def inference(self):
        with torch.no_grad():
            outputs = self.model(self.x)
        print("result: " + class_names[torch.argmax(outputs, 1)[0]])

    def prof(self, depth=-1):
        with pytorchtool.Profile(self.model, use_cuda=self.use_gpu,
                depth=depth) as prof:
            outputs = self.model(self.x)
            print("result: " + class_names[torch.argmax(outputs, 1)[0]])

        if not os.path.exists("./parameters/" + self.model_name):
            os.makedirs("./parameters/" + self.model_name)
        if self.use_gpu:
            prof.printCsv("./parameters/" + self.model_name + "/gpuPart.csv")
        else:
            prof.printCsv("./parameters/" + self.model_name + "/cpuPart.csv")


if __name__ == "__main__":
    torch.randn(4).to(0)

    name = "res"
    start_init = time.time()
    m = model(name, use_gpu=True)
    print("模型结构初始化时间: ", time.time() - start_init)
    start_load = time.time()
    m.load_weight()
    print("模型参数加载时间: ", time.time() - start_load)

    m.inference()

    doPrepare = False
    doProf = True
    doInference = False
    doPartition = False
    doPartition2 = False

    if doPrepare:
        m.save_layers(depth=m.depth)
    elif doProf:
        m.prof(depth=m.depth)
        m.prof(depth=m.depth)
    elif doInference:
        start = time.time()
        m.inference()
        print("推理时间", time.time() - start)
    elif doPartition:
        '''
        使用Alexnet进行了切分测试
        '''
        cModel = pytorchtool.Surgery(m.model, 0, depth=m.depth)
        cModel.setLayerState({"input": 1, "features.0": 2, "features.1": 2, "features.2": 2, "features.3": 2,
                            "features.4": 2, "features.5": 2, "features.6": 2, "features.7": 2,
                            "features.8": 2, "features.9": 2, "features.10": 2, "features.11": 2,
                            "features.12": 2, "avgpool": 2, "classifier.0": 2, "classifier.1": 2,
                            "classifier.2": 2, "classifier.3": 2, "classifier.4": 2, "classifier.5": 2,
                            "classifier.6": 2, 'flatten': 2})
        cModel.clearMiddleResult()
        cModel(m.x)
        cModel.recover() # 恢复m的forward函数，避免sModel对同一个模型嵌套修改
        print(cModel.getMiddleResult())

        sModel = pytorchtool.Surgery(m.model, 2, depth=m.depth)
        sModel.setLayerState({"input": 1, "features.0": 2, "features.1": 2, "features.2": 2, "features.3": 2,
                            "features.4": 2, "features.5": 2, "features.6": 2, "features.7": 2,
                            "features.8": 2, "features.9": 2, "features.10": 2, "features.11": 2,
                            "features.12": 2, "avgpool": 2, "classifier.0": 2, "classifier.1": 2,
                            "classifier.2": 2, "classifier.3": 2, "classifier.4": 2, "classifier.5": 2,
                            "classifier.6": 2, 'flatten': 2})
        '''
        这里使用随机生成的相同size的数据代替原始输入数据，
        实际使用时若将计算全部卸载到了服务端，则需要传入原始数据
        '''
        sModel.setMiddleResult(cModel.getMiddleResult())
        outputs = sModel(torch.rand(224, 224).unsqueeze_(0))
        print("result: " + class_names[torch.argmax(outputs, 1)[0]])
    elif doPartition2:
        '''
        使用Alexnet进行了切分测试
        '''
        cModel = pytorchtool.Surgery(m.model, 0, depth=m.depth)
        cModel.setLayerState({"input": 1, "features.0": 2, "features.1": 2, "features.2": 2, "features.3": 2,
                            "features.4": 2, "features.5": 2, "features.6": 2, "features.7": 2,
                            "features.8": 2, "features.9": 2, "features.10": 2, "features.11": 2,
                            "features.12": 2, "avgpool": 2, "classifier.0": 2, "classifier.1": 2,
                            "classifier.2": 2, "classifier.3": 2, "classifier.4": 2, "classifier.5": 2,
                            "classifier.6": 2, 'flatten': 2})
        cModel.clearMiddleResult()
        cModel(m.x)
        cModel.recover() # 恢复m的forward函数，避免sModel对同一个模型嵌套修改
        print(cModel.getMiddleResult())

        sModel = pytorchtool.Surgery2('alex', './parameters/alexnet/dag')
        for k, v in {"input": 1, "features.0": 2, "features.1": 2, "features.2": 2, "features.3": 2,
                            "features.4": 2, "features.5": 2, "features.6": 2, "features.7": 2,
                            "features.8": 2, "features.9": 2, "features.10": 2, "features.11": 2,
                            "features.12": 2, "avgpool": 2, "classifier.0": 2, "classifier.1": 2,
                            "classifier.2": 2, "classifier.3": 2, "classifier.4": 2, "classifier.5": 2,
                            "classifier.6": 2, 'flatten': 2}.items():
            if v == 2:
                sModel.loadLayer(k)

        start = time.time()
        outputs = sModel.inferencePart(cModel.getMiddleResult())
        print("服务端时间", time.time() - start)
        print("result: " + class_names[torch.argmax(outputs, 1)[0]])