pytorch 计算模型的GFlops和total params的方法
2024-04-11 20:32:51
建议使用两个第三方库torchstat和torchsummary
- 自己计算使用parameters
方式一
使用model.paramters
num_params = 0
for param in model.parameters():num_params += param.numel()
print(num_params)
print(f"params : {num_params/1e6}M")
# print(model.net)方式二
使用model.named_parameters()
num_params = 0
for name, param in model.named_parameters():# print(name,param.numel())num_params += param.numel()
print(num_params)
print(f"params : {num_params/1e6} M")
方式三
def get_parameter_number(net):total_num = sum(p.numel() for p in net.parameters())trainable_num = sum(p.numel() for p in net.parameters() if p.requires_grad)return {'Total': total_num, 'Trainable': trainable_num}print(get_parameter_number(model))- torchstat
使用pip install torchstat就可以安装torchstat了
pip install torchstat
from torchstat import stat
from torchvision.models.resnet import resnet34
model = resnet34()
stat(model, (3, 224, 224))[MAdd]: AdaptiveAvgPool2d is not supported!
[Flops]: AdaptiveAvgPool2d is not supported!
[Memory]: AdaptiveAvgPool2d is not supported!module name input shape output shape params memory(MB) MAdd Flops MemRead(B) MemWrite(B) duration[%] MemR+W(B)
0 conv1 3 224 224 64 112 112 9408.0 3.06 235,225,088.0 118,013,952.0 639744.0 3211264.0 8.76% 3851008.0
1 bn1 64 112 112 64 112 112 128.0 3.06 3,211,264.0 1,605,632.0 3211776.0 3211264.0 1.26% 6423040.0
2 relu 64 112 112 64 112 112 0.0 3.06 802,816.0 802,816.0 3211264.0 3211264.0 0.43% 6422528.0
3 maxpool 64 112 112 64 56 56 0.0 0.77 1,605,632.0 802,816.0 3211264.0 802816.0 2.00% 4014080.0
4 layer1.0.conv1 64 56 56 64 56 56 36864.0 0.77 231,010,304.0 115,605,504.0 950272.0 802816.0 3.26% 1753088.0
5 layer1.0.bn1 64 56 56 64 56 56 128.0 0.77 802,816.0 401,408.0 803328.0 802816.0 0.41% 1606144.0
6 layer1.0.relu 64 56 56 64 56 56 0.0 0.77 200,704.0 200,704.0 802816.0 802816.0 0.07% 1605632.0
7 layer1.0.conv2 64 56 56 64 56 56 36864.0 0.77 231,010,304.0 115,605,504.0 950272.0 802816.0 2.26% 1753088.0
8 layer1.0.bn2 64 56 56 64 56 56 128.0 0.77 802,816.0 401,408.0 803328.0 802816.0 0.40% 1606144.0
9 layer1.1.conv1 64 56 56 64 56 56 36864.0 0.77 231,010,304.0 115,605,504.0 950272.0 802816.0 2.43% 1753088.0
10 layer1.1.bn1 64 56 56 64 56 56 128.0 0.77 802,816.0 401,408.0 803328.0 802816.0 0.35% 1606144.0
11 layer1.1.relu 64 56 56 64 56 56 0.0 0.77 200,704.0 200,704.0 802816.0 802816.0 0.06% 1605632.0
12 layer1.1.conv2 64 56 56 64 56 56 36864.0 0.77 231,010,304.0 115,605,504.0 950272.0 802816.0 2.28% 1753088.0
13 layer1.1.bn2 64 56 56 64 56 56 128.0 0.77 802,816.0 401,408.0 803328.0 802816.0 0.38% 1606144.0
14 layer1.2.conv1 64 56 56 64 56 56 36864.0 0.77 231,010,304.0 115,605,504.0 950272.0 802816.0 2.38% 1753088.0
15 layer1.2.bn1 64 56 56 64 56 56 128.0 0.77 802,816.0 401,408.0 803328.0 802816.0 0.32% 1606144.0
16 layer1.2.relu 64 56 56 64 56 56 0.0 0.77 200,704.0 200,704.0 802816.0 802816.0 0.06% 1605632.0
17 layer1.2.conv2 64 56 56 64 56 56 36864.0 0.77 231,010,304.0 115,605,504.0 950272.0 802816.0 2.28% 1753088.0
18 layer1.2.bn2 64 56 56 64 56 56 128.0 0.77 802,816.0 401,408.0 803328.0 802816.0 0.41% 1606144.0
19 layer2.0.conv1 64 56 56 128 28 28 73728.0 0.38 115,505,152.0 57,802,752.0 1097728.0 401408.0 2.26% 1499136.0
20 layer2.0.bn1 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.17% 803840.0
21 layer2.0.relu 128 28 28 128 28 28 0.0 0.38 100,352.0 100,352.0 401408.0 401408.0 0.06% 802816.0
22 layer2.0.conv2 128 28 28 128 28 28 147456.0 0.38 231,110,656.0 115,605,504.0 991232.0 401408.0 2.46% 1392640.0
23 layer2.0.bn2 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.17% 803840.0
24 layer2.0.downsample.0 64 56 56 128 28 28 8192.0 0.38 12,744,704.0 6,422,528.0 835584.0 401408.0 1.67% 1236992.0
25 layer2.0.downsample.1 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.17% 803840.0
26 layer2.1.conv1 128 28 28 128 28 28 147456.0 0.38 231,110,656.0 115,605,504.0 991232.0 401408.0 1.77% 1392640.0
27 layer2.1.bn1 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.17% 803840.0
28 layer2.1.relu 128 28 28 128 28 28 0.0 0.38 100,352.0 100,352.0 401408.0 401408.0 0.06% 802816.0
29 layer2.1.conv2 128 28 28 128 28 28 147456.0 0.38 231,110,656.0 115,605,504.0 991232.0 401408.0 1.87% 1392640.0
30 layer2.1.bn2 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.19% 803840.0
31 layer2.2.conv1 128 28 28 128 28 28 147456.0 0.38 231,110,656.0 115,605,504.0 991232.0 401408.0 1.63% 1392640.0
32 layer2.2.bn1 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.18% 803840.0
33 layer2.2.relu 128 28 28 128 28 28 0.0 0.38 100,352.0 100,352.0 401408.0 401408.0 0.06% 802816.0
34 layer2.2.conv2 128 28 28 128 28 28 147456.0 0.38 231,110,656.0 115,605,504.0 991232.0 401408.0 1.68% 1392640.0
35 layer2.2.bn2 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.17% 803840.0
36 layer2.3.conv1 128 28 28 128 28 28 147456.0 0.38 231,110,656.0 115,605,504.0 991232.0 401408.0 1.66% 1392640.0
37 layer2.3.bn1 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.16% 803840.0
38 layer2.3.relu 128 28 28 128 28 28 0.0 0.38 100,352.0 100,352.0 401408.0 401408.0 0.06% 802816.0
39 layer2.3.conv2 128 28 28 128 28 28 147456.0 0.38 231,110,656.0 115,605,504.0 991232.0 401408.0 1.70% 1392640.0
40 layer2.3.bn2 128 28 28 128 28 28 256.0 0.38 401,408.0 200,704.0 402432.0 401408.0 0.16% 803840.0
41 layer3.0.conv1 128 28 28 256 14 14 294912.0 0.19 115,555,328.0 57,802,752.0 1581056.0 200704.0 2.17% 1781760.0
42 layer3.0.bn1 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.14% 403456.0
43 layer3.0.relu 256 14 14 256 14 14 0.0 0.19 50,176.0 50,176.0 200704.0 200704.0 0.10% 401408.0
44 layer3.0.conv2 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 2.64% 2760704.0
45 layer3.0.bn2 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.12% 403456.0
46 layer3.0.downsample.0 128 28 28 256 14 14 32768.0 0.19 12,794,880.0 6,422,528.0 532480.0 200704.0 1.23% 733184.0
47 layer3.0.downsample.1 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.13% 403456.0
48 layer3.1.conv1 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 1.68% 2760704.0
49 layer3.1.bn1 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.12% 403456.0
50 layer3.1.relu 256 14 14 256 14 14 0.0 0.19 50,176.0 50,176.0 200704.0 200704.0 0.06% 401408.0
51 layer3.1.conv2 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 1.64% 2760704.0
52 layer3.1.bn2 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.13% 403456.0
53 layer3.2.conv1 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 1.64% 2760704.0
54 layer3.2.bn1 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.12% 403456.0
55 layer3.2.relu 256 14 14 256 14 14 0.0 0.19 50,176.0 50,176.0 200704.0 200704.0 0.05% 401408.0
56 layer3.2.conv2 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 1.65% 2760704.0
57 layer3.2.bn2 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.12% 403456.0
58 layer3.3.conv1 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 1.73% 2760704.0
59 layer3.3.bn1 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.13% 403456.0
60 layer3.3.relu 256 14 14 256 14 14 0.0 0.19 50,176.0 50,176.0 200704.0 200704.0 0.06% 401408.0
61 layer3.3.conv2 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 2.13% 2760704.0
62 layer3.3.bn2 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.14% 403456.0
63 layer3.4.conv1 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 1.67% 2760704.0
64 layer3.4.bn1 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.14% 403456.0
65 layer3.4.relu 256 14 14 256 14 14 0.0 0.19 50,176.0 50,176.0 200704.0 200704.0 0.05% 401408.0
66 layer3.4.conv2 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 1.68% 2760704.0
67 layer3.4.bn2 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.12% 403456.0
68 layer3.5.conv1 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 1.95% 2760704.0
69 layer3.5.bn1 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.14% 403456.0
70 layer3.5.relu 256 14 14 256 14 14 0.0 0.19 50,176.0 50,176.0 200704.0 200704.0 0.06% 401408.0
71 layer3.5.conv2 256 14 14 256 14 14 589824.0 0.19 231,160,832.0 115,605,504.0 2560000.0 200704.0 2.01% 2760704.0
72 layer3.5.bn2 256 14 14 256 14 14 512.0 0.19 200,704.0 100,352.0 202752.0 200704.0 0.14% 403456.0
73 layer4.0.conv1 256 14 14 512 7 7 1179648.0 0.10 115,580,416.0 57,802,752.0 4919296.0 100352.0 2.81% 5019648.0
74 layer4.0.bn1 512 7 7 512 7 7 1024.0 0.10 100,352.0 50,176.0 104448.0 100352.0 0.13% 204800.0
75 layer4.0.relu 512 7 7 512 7 7 0.0 0.10 25,088.0 25,088.0 100352.0 100352.0 0.04% 200704.0
76 layer4.0.conv2 512 7 7 512 7 7 2359296.0 0.10 231,185,920.0 115,605,504.0 9537536.0 100352.0 6.47% 9637888.0
77 layer4.0.bn2 512 7 7 512 7 7 1024.0 0.10 100,352.0 50,176.0 104448.0 100352.0 0.13% 204800.0
78 layer4.0.downsample.0 256 14 14 512 7 7 131072.0 0.10 12,819,968.0 6,422,528.0 724992.0 100352.0 1.46% 825344.0
79 layer4.0.downsample.1 512 7 7 512 7 7 1024.0 0.10 100,352.0 50,176.0 104448.0 100352.0 0.14% 204800.0
80 layer4.1.conv1 512 7 7 512 7 7 2359296.0 0.10 231,185,920.0 115,605,504.0 9537536.0 100352.0 3.55% 9637888.0
81 layer4.1.bn1 512 7 7 512 7 7 1024.0 0.10 100,352.0 50,176.0 104448.0 100352.0 0.13% 204800.0
82 layer4.1.relu 512 7 7 512 7 7 0.0 0.10 25,088.0 25,088.0 100352.0 100352.0 0.03% 200704.0
83 layer4.1.conv2 512 7 7 512 7 7 2359296.0 0.10 231,185,920.0 115,605,504.0 9537536.0 100352.0 3.06% 9637888.0
84 layer4.1.bn2 512 7 7 512 7 7 1024.0 0.10 100,352.0 50,176.0 104448.0 100352.0 0.13% 204800.0
85 layer4.2.conv1 512 7 7 512 7 7 2359296.0 0.10 231,185,920.0 115,605,504.0 9537536.0 100352.0 2.94% 9637888.0
86 layer4.2.bn1 512 7 7 512 7 7 1024.0 0.10 100,352.0 50,176.0 104448.0 100352.0 0.14% 204800.0
87 layer4.2.relu 512 7 7 512 7 7 0.0 0.10 25,088.0 25,088.0 100352.0 100352.0 0.03% 200704.0
88 layer4.2.conv2 512 7 7 512 7 7 2359296.0 0.10 231,185,920.0 115,605,504.0 9537536.0 100352.0 3.02% 9637888.0
89 layer4.2.bn2 512 7 7 512 7 7 1024.0 0.10 100,352.0 50,176.0 104448.0 100352.0 0.12% 204800.0
90 avgpool 512 7 7 512 1 1 0.0 0.00 0.0 0.0 0.0 0.0 0.38% 0.0
91 fc 512 1000 513000.0 0.00 1,023,000.0 512,000.0 2054048.0 4000.0 1.14% 2058048.0
total 21797672.0 37.62 7,342,524,440.0 3,674,223,104.0 2054048.0 4000.0 100.00% 167277632.0
=================================================================================================================================================================
Total params: 21,797,672
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
Total memory: 37.62MB
Total MAdd: 7.34GMAdd
Total Flops: 3.67GFlops
Total MemR+W: 159.53MB
- 使用torchsummary
安装，使用pip 安装
pip install torchsummary
from torchvision.models.resnet import resnet34
from torchsummary import summary
import torchmodel = resnet34(pretrained=False).eval().cuda()
summary(model, input_size=(3, 224, 224), batch_size=-1)
----------------------------------------------------------------Layer (type) Output Shape Param #
================================================================Conv2d-1 [-1, 64, 112, 112] 9,408BatchNorm2d-2 [-1, 64, 112, 112] 128ReLU-3 [-1, 64, 112, 112] 0MaxPool2d-4 [-1, 64, 56, 56] 0Conv2d-5 [-1, 64, 56, 56] 36,864BatchNorm2d-6 [-1, 64, 56, 56] 128ReLU-7 [-1, 64, 56, 56] 0Conv2d-8 [-1, 64, 56, 56] 36,864BatchNorm2d-9 [-1, 64, 56, 56] 128ReLU-10 [-1, 64, 56, 56] 0BasicBlock-11 [-1, 64, 56, 56] 0Conv2d-12 [-1, 64, 56, 56] 36,864BatchNorm2d-13 [-1, 64, 56, 56] 128ReLU-14 [-1, 64, 56, 56] 0Conv2d-15 [-1, 64, 56, 56] 36,864BatchNorm2d-16 [-1, 64, 56, 56] 128ReLU-17 [-1, 64, 56, 56] 0BasicBlock-18 [-1, 64, 56, 56] 0Conv2d-19 [-1, 64, 56, 56] 36,864BatchNorm2d-20 [-1, 64, 56, 56] 128ReLU-21 [-1, 64, 56, 56] 0Conv2d-22 [-1, 64, 56, 56] 36,864BatchNorm2d-23 [-1, 64, 56, 56] 128ReLU-24 [-1, 64, 56, 56] 0BasicBlock-25 [-1, 64, 56, 56] 0Conv2d-26 [-1, 128, 28, 28] 73,728BatchNorm2d-27 [-1, 128, 28, 28] 256ReLU-28 [-1, 128, 28, 28] 0Conv2d-29 [-1, 128, 28, 28] 147,456BatchNorm2d-30 [-1, 128, 28, 28] 256Conv2d-31 [-1, 128, 28, 28] 8,192BatchNorm2d-32 [-1, 128, 28, 28] 256ReLU-33 [-1, 128, 28, 28] 0BasicBlock-34 [-1, 128, 28, 28] 0Conv2d-35 [-1, 128, 28, 28] 147,456BatchNorm2d-36 [-1, 128, 28, 28] 256ReLU-37 [-1, 128, 28, 28] 0Conv2d-38 [-1, 128, 28, 28] 147,456BatchNorm2d-39 [-1, 128, 28, 28] 256ReLU-40 [-1, 128, 28, 28] 0BasicBlock-41 [-1, 128, 28, 28] 0Conv2d-42 [-1, 128, 28, 28] 147,456BatchNorm2d-43 [-1, 128, 28, 28] 256ReLU-44 [-1, 128, 28, 28] 0Conv2d-45 [-1, 128, 28, 28] 147,456BatchNorm2d-46 [-1, 128, 28, 28] 256ReLU-47 [-1, 128, 28, 28] 0BasicBlock-48 [-1, 128, 28, 28] 0Conv2d-49 [-1, 128, 28, 28] 147,456BatchNorm2d-50 [-1, 128, 28, 28] 256ReLU-51 [-1, 128, 28, 28] 0Conv2d-52 [-1, 128, 28, 28] 147,456BatchNorm2d-53 [-1, 128, 28, 28] 256ReLU-54 [-1, 128, 28, 28] 0BasicBlock-55 [-1, 128, 28, 28] 0Conv2d-56 [-1, 256, 14, 14] 294,912BatchNorm2d-57 [-1, 256, 14, 14] 512ReLU-58 [-1, 256, 14, 14] 0Conv2d-59 [-1, 256, 14, 14] 589,824BatchNorm2d-60 [-1, 256, 14, 14] 512Conv2d-61 [-1, 256, 14, 14] 32,768BatchNorm2d-62 [-1, 256, 14, 14] 512ReLU-63 [-1, 256, 14, 14] 0BasicBlock-64 [-1, 256, 14, 14] 0Conv2d-65 [-1, 256, 14, 14] 589,824BatchNorm2d-66 [-1, 256, 14, 14] 512ReLU-67 [-1, 256, 14, 14] 0Conv2d-68 [-1, 256, 14, 14] 589,824BatchNorm2d-69 [-1, 256, 14, 14] 512ReLU-70 [-1, 256, 14, 14] 0BasicBlock-71 [-1, 256, 14, 14] 0Conv2d-72 [-1, 256, 14, 14] 589,824BatchNorm2d-73 [-1, 256, 14, 14] 512ReLU-74 [-1, 256, 14, 14] 0Conv2d-75 [-1, 256, 14, 14] 589,824BatchNorm2d-76 [-1, 256, 14, 14] 512ReLU-77 [-1, 256, 14, 14] 0BasicBlock-78 [-1, 256, 14, 14] 0Conv2d-79 [-1, 256, 14, 14] 589,824BatchNorm2d-80 [-1, 256, 14, 14] 512ReLU-81 [-1, 256, 14, 14] 0Conv2d-82 [-1, 256, 14, 14] 589,824BatchNorm2d-83 [-1, 256, 14, 14] 512ReLU-84 [-1, 256, 14, 14] 0BasicBlock-85 [-1, 256, 14, 14] 0Conv2d-86 [-1, 256, 14, 14] 589,824BatchNorm2d-87 [-1, 256, 14, 14] 512ReLU-88 [-1, 256, 14, 14] 0Conv2d-89 [-1, 256, 14, 14] 589,824BatchNorm2d-90 [-1, 256, 14, 14] 512ReLU-91 [-1, 256, 14, 14] 0BasicBlock-92 [-1, 256, 14, 14] 0Conv2d-93 [-1, 256, 14, 14] 589,824BatchNorm2d-94 [-1, 256, 14, 14] 512ReLU-95 [-1, 256, 14, 14] 0Conv2d-96 [-1, 256, 14, 14] 589,824BatchNorm2d-97 [-1, 256, 14, 14] 512ReLU-98 [-1, 256, 14, 14] 0BasicBlock-99 [-1, 256, 14, 14] 0Conv2d-100 [-1, 512, 7, 7] 1,179,648BatchNorm2d-101 [-1, 512, 7, 7] 1,024ReLU-102 [-1, 512, 7, 7] 0Conv2d-103 [-1, 512, 7, 7] 2,359,296BatchNorm2d-104 [-1, 512, 7, 7] 1,024Conv2d-105 [-1, 512, 7, 7] 131,072BatchNorm2d-106 [-1, 512, 7, 7] 1,024ReLU-107 [-1, 512, 7, 7] 0BasicBlock-108 [-1, 512, 7, 7] 0Conv2d-109 [-1, 512, 7, 7] 2,359,296BatchNorm2d-110 [-1, 512, 7, 7] 1,024ReLU-111 [-1, 512, 7, 7] 0Conv2d-112 [-1, 512, 7, 7] 2,359,296BatchNorm2d-113 [-1, 512, 7, 7] 1,024ReLU-114 [-1, 512, 7, 7] 0BasicBlock-115 [-1, 512, 7, 7] 0Conv2d-116 [-1, 512, 7, 7] 2,359,296BatchNorm2d-117 [-1, 512, 7, 7] 1,024ReLU-118 [-1, 512, 7, 7] 0Conv2d-119 [-1, 512, 7, 7] 2,359,296BatchNorm2d-120 [-1, 512, 7, 7] 1,024ReLU-121 [-1, 512, 7, 7] 0BasicBlock-122 [-1, 512, 7, 7] 0
AdaptiveAvgPool2d-123 [-1, 512, 1, 1] 0Linear-124 [-1, 1000] 513,000
================================================================
Total params: 21,797,672
Trainable params: 21,797,672
Non-trainable params: 0
----------------------------------------------------------------
Input size (MB): 0.57
Forward/backward pass size (MB): 96.29
Params size (MB): 83.15
Estimated Total Size (MB): 180.01
----------------------------------------------------------------
pytorch 计算模型的GFlops和total params的方法相关推荐
- pytorch计算模型参数量
1. 安装 thop 1.1 常规安装 pip install thop 1.2 若上述安装方式错误,可以参考以下方式: pip install thop-i http://pypi.douban.c ...
- pytorch 获取模型参数_Pytorch获取模型参数情况的方法
分享人工智能技术干货,专注深度学习与计算机视觉领域! 相较于Tensorflow,Pytorch一开始就是以动态图构建神经网络图的,其获取模型参数的方法也比较容易,既可以根据其内建接口自己写代码获取模 ...
- 《融智学进阶文集》01：间接计算模型和间接形式化方法
<融智学进阶文集>01: 间接计算模型和间接形式化方法 01-间接计算模型和间接形式化方法_邹晓辉.pdf 怎么采用融智学七遍通方法熟悉原创文本? 对照阅读 摘要: 本文旨在:从人机交互界 ...
- 【深度学习】深度学习中模型计算量(FLOPs)和参数量(Params)等的理解以及四种在python应用的计算方法总结
接下来要分别概述以下内容: 1 首先什么是参数量,什么是计算量 2 如何计算 参数量,如何统计 计算量 3 换算参数量,把他换算成我们常用的单位,比如:mb 4 对于各个经典网络,论述他们是计算量大还 ...
- Pytorch 获取模型 Params/FLOPS
Note! 查看 model 参数值 model.state_dict() 1.自定义 Params Pytorch依据其内建接口自己写代码获取模型参数情况,大家可以参考Pytorch提供的model ...
- pytorch模型参数信息 计算模型的FLOPs
参考链接:https://blog.csdn.net/tsq292978891/article/details/87918244 打印模型参数信息 在python3环境下安装torchsummary ...
- pytorch计算FLOPs
转自:pytorch计算FLOPs - 简书 1. 引言 其实模型的参数量好算,但浮点运算数并不好确定,我们一般也就根据参数量直接估计计算量了.但是像卷积之类的运算,它的参数量比较小,但是运算量非常大 ...
- 神经网络的计算量(FLOPs)、参数量(Params)、推理时间（FPS）的定义及实现方法
目录 1. 定义 2. 实现方法 2.1. 计算参数量 2.2. 计算参数量和FLOPs 2.3. 计算推理时间(FPS) 3. 数据大小对参数量和FLOPs的影响 4. 参数量和FLOPs对于硬件要 ...
- pytorch保存模型pth_Day159:模型的保存与加载
网络结构和参数可以分开的保存和加载,因此,pytorch保存模型有两种方法: 保存 整个模型 (结构+参数) 只保存模型参数(官方推荐) # 保存整个网络torch.save(model, check ...
- c++list遍历_小白学PyTorch | 6 模型的构建访问遍历存储（附代码）
关注一下不迷路哦~喜欢的点个星标吧~<> 小白学PyTorch | 5 torchvision预训练模型与数据集全览 小白学PyTorch | 4 构建模型三要素与权重初始化 小白学PyT ...
最新文章
- 关不关机 扫地机器人_【小米智能家居】米家扫拖机器人，模拟人工来回擦拖地！...
- SSH远程登录解析（linux）
- VTK：提取可见细胞用法实战
- php curl 下载图片,CURL实现下载远程图片并保存到本地
- 使用StringWriter和StringReader的好处
- 火狐SEO插件：查询网站收录与外链的火狐油猴脚本工具
- Mac上的一位数密码你知道吗
- ECS上配置FTP Filezilla
- 进程以及状态 进程的创建
- 基于近邻用户协同过滤算法的音乐推荐系统
- 项目开发：网上书店（详细的开发流程记录）----注册登录功能，通过邮件验证
- JS 延时函数 setTimeout 或者 rxjs 写法
- 新闻资讯|iPad mini 概念想象重新设计，配备更大的8.4英寸Liquid Retina显示屏
- “马赛克”真能去除了？老司机狂喜！这一神器一键去除！
- Java代码安装maven jar_Java中Maven项目导出jar包配置的示例代码
- 你什么也无法告诉别人
- 桌面widget详解（四）——桌面音乐播放器（实战）
- ieee 802.3学习笔记-MII
- 笔试真题：100颗糖果，甲乙轮流从糖果盒中取出糖果，每次可取出2、4或6颗，若取得最后糖果的玩家为最终胜者，若甲先取z则（甲获胜，乙获胜，平局，不确定）
- Chain of Responsibility模式——读书笔记