陈云pytorch学习笔记_用50行代码搭建ResNet
import torch as t
import torch.nn as nn
import torch.nn.functional as F
from torchvision import models
# 残差快 残差网络公式 a^[L+] = g(a^[L]+z^[L+])
class ResidualBlock(nn.Module):
def __init__(self, inchannel, outchannel, stride=, shortcut=None): #shortcut=None对应图中跨层连接的实线,对应残差网络公式 a^[L+] = g(a^[L]+z^[L+])，否则对应当
# 通道数变化后第一个残差块的虚线，此时对应的残差公式为a^[L+] = g(z^[L+]+z^[L+])
nn.Module.__init__(self)
self.left = nn.Sequential(#得到z^[L+]
nn.Conv2d(inchannel, outchannel, , stride, , bias=False),
nn.BatchNorm2d(outchannel),
nn.ReLU(inplace= True),
nn.Conv2d(outchannel, outchannel, , , , bias=False),
nn.BatchNorm2d(outchannel))
self.right = shortcut#决定是跨层连接的是实线还是虚线
def forward(self, x):
out = self.left(x)
residual = x if self.right is None else self.right(x)
out += residual
return F.relu(out) #a^[L+] = g(a^[L]+z^[L+])
# ResNet34
class ResNet(nn.Module):
def __init__(self, num_classes=):
nn.Module.__init__(self)
# 前几层图像转换(网络输入部分)
self.pre = nn.Sequential(#对应图中开始残差处理之前的部分
nn.Conv2d(, , kernel_size=, stride=, padding=, bias=False),
nn.BatchNorm2d(),
nn.ReLU(inplace=True),
nn.MaxPool2d(, , )
)
# 中间卷积部分
self.layer1 = self._make_layer(, , )
self.layer2 = self._make_layer(, , , stride=)#stride=2代表每一个残差快的第一个层的2/
self.layer3 = self._make_layer(, , , stride=)
self.layer4 = self._make_layer(, , , stride=)
# 平均池化
self.avgpool = nn.AvgPool2d(, stride=)
# 分类用的全连接
self.fc = nn.Linear(, )
def _make_layer(self, inchannel, outchannel, block_num, stride=):
# 使得输入输出通道数调整为一致。比如第二个layer时，第一个残差快输入为64，输出为128
shortcut = nn.Sequential(#对应着每类相同通道数的残差快的第一个跨层直线是虚线
nn.Conv2d(inchannel, outchannel, , stride, bias=False),
nn.BatchNorm2d(outchannel))
layers = []
layers.append(ResidualBlock(inchannel, outchannel, stride, shortcut))
for i in range(, block_num):
layers.append(ResidualBlock(outchannel, outchannel))
return nn.Sequential(*layers)
def forward(self, x):
x = self.pre(x)
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.layer4(x)
x = self.avgpool(x)
x = x.view(x.size(), -)#torch.Size([, ])
return self.fc(x) model = ResNet()
input = t.autograd.Variable(t.randn(, , , ))
o = model(input)
print(o)
model = models.resnet34()#调用工具包实线残差网络
o1 = model(input)
print(o1)
output
D:\anaconda\anaconda\pythonw.exe D:/Code/Python/pytorch入门与实践/第四章_神经网络工具箱nn/搭建ResNet.py
tensor([[-3.7385e-01, -2.2010e-01, 8.9218e-01, -6.6067e-01, -5.8422e-01,
7.6649e-01, -8.0401e-01, -3.0225e-01, 2.4314e+00, 9.1019e-01,
-1.1270e+00, -2.9847e-01, -6.0022e-01, -3.5480e-01, 3.5396e-01,
3.6958e-01, -5.9464e-01, -4.5049e-01, 7.3531e-01, -1.1082e+00,
-2.9160e-01, 4.7690e-01, -5.0259e-01, 4.1628e-01, -9.3588e-01,
2.4529e-01, 1.2500e+00, -1.6038e-01, -3.3023e-01, 3.6957e-01,
4.5195e-01, -6.0984e-01, 8.9558e-02, 4.2407e-01, 1.2888e+00,
-5.3017e-01, -5.0509e-01, 1.3775e+00, 2.0299e-01, 2.7299e-01,
-8.6381e-02, 8.3670e-01, 1.9371e-01, 8.4775e-01, 4.0002e-01,
8.1064e-01, -1.0556e+00, 7.7371e-01, 6.8891e-01, -1.0209e+00,
-3.1227e-01, 3.0977e-02, -2.9849e-01, 1.0534e+00, -1.8499e-01,
-8.6205e-01, -3.4020e-01, 4.2196e-01, -1.0207e-01, -1.2846e+00,
3.1905e-01, 9.9212e-01, 6.0208e-01, 6.2960e-01, 4.1285e-01,
-1.1597e+00, -2.5292e-01, 7.0145e-01, 4.8221e-01, 6.6790e-01,
2.8644e-01, -2.6554e-01, 3.4255e-01, 3.8341e-01, -4.2387e-01,
-7.5981e-01, 2.3203e-01, -3.8434e-01, -6.9346e-01, -5.4996e-01,
2.7865e-01, -3.9543e-01, -7.8868e-01, -6.5345e-01, 2.9946e-01,
5.2693e-01, -4.5380e-01, -2.2429e-01, -1.4129e+00, -3.7963e-01,
-2.2408e-01, -6.9917e-02, 1.7414e-01, -2.7821e-01, -6.4848e-01,
-4.3716e-01, -3.5371e-01, -3.6100e-01, -5.7401e-01, 3.7754e-01,
-4.1583e-02, -3.4307e-01, 5.9179e-01, 2.7279e-02, -2.6988e-01,
-4.7790e-01, -3.5140e-01, 5.1556e-01, 8.0434e-01, -4.4143e-01,
6.0849e-01, 3.3159e-01, -7.6929e-03, 1.0759e-01, 2.6402e-01,
2.3914e-01, -6.5949e-01, -9.0380e-01, -5.7449e-01, 8.1698e-01,
9.0535e-01, 3.7668e-01, 3.2937e-01, 4.5524e-01, 1.6086e-01,
-2.8713e-01, 1.7160e-01, 3.5057e-01, 7.0938e-01, -6.3579e-02,
-3.9463e-01, 2.6736e-01, -4.4593e-01, 1.0601e+00, -3.6988e-01,
-6.2878e-01, 3.7628e-01, 5.3490e-01, -3.2025e-01, -6.2648e-01,
-5.2117e-02, -4.0097e-01, -1.1775e+00, 1.2687e+00, 1.1808e+00,
3.4300e-01, -2.3935e-01, -7.8519e-01, 3.3952e-01, -2.1779e-01,
-4.9251e-01, -4.1354e-01, -7.1647e-01, 1.1502e+00, -9.0239e-01,
-1.8571e-01, 8.7283e-01, 5.6701e-01, 9.5695e-02, -2.6622e-01,
3.3122e-02, 5.8339e-01, 6.4253e-01, -1.2866e-01, 1.9386e+00,
3.0843e-01, 1.7281e-01, 1.3516e-01, -7.3507e-02, -3.3128e-01,
-6.3045e-02, -1.6130e-01, -6.2078e-01, 1.0369e+00, 7.4816e-01,
4.3222e-01, -1.2471e+00, -2.8628e-01, -2.2325e-01, -1.3061e+00,
6.0621e-01, 1.2517e+00, 7.9576e-01, 1.3829e-01, 3.0933e-01,
-5.7864e-01, -1.3680e-01, 4.2718e-01, 6.0374e-01, -6.5616e-01,
8.8827e-01, 6.1121e-01, 9.2531e-01, -9.0994e-01, -1.6550e+00,
-9.5535e-01, -1.3156e+00, 2.3245e-01, -4.6053e-01, 1.9782e-01,
-7.3612e-01, 7.3810e-01, -6.6007e-02, -5.0354e-01, 5.7257e-01,
-8.2178e-02, -1.0175e+00, -7.8140e-01, -8.3596e-02, 4.3341e-01,
6.1036e-01, -6.0388e-01, 3.3036e-01, 3.2923e-01, 1.2033e+00,
-6.1371e-01, -8.7145e-01, -7.0251e-02, -1.9632e-01, -4.0972e-01,
-6.5015e-01, -1.1036e+00, 4.5884e-01, -7.8906e-01, 9.7192e-01,
7.7442e-01, 3.4869e-01, 9.7635e-02, -9.9016e-01, 7.6778e-01,
3.5343e-01, 1.1142e-01, -1.4715e-01, -3.1201e-01, -4.6759e-01,
4.2290e-01, 2.9731e-01, -5.6528e-01, -4.6112e-01, -9.2171e-03,
2.7790e-01, 2.2434e-01, 8.3167e-01, 6.0836e-01, 7.9597e-01,
8.8949e-01, -5.5800e-01, -5.8002e-01, 2.3448e-02, -3.7334e-01,
-2.2329e-01, -1.1076e+00, -3.0460e-01, 1.4154e-02, 3.7740e-04,
9.9988e-02, -7.1763e-01, -2.9103e-01, 3.7885e-01, 4.5475e-01,
-8.8300e-01, -5.9084e-01, 1.0630e-01, -7.6122e-01, 5.6615e-01,
1.5967e-02, -1.3541e+00, -2.3975e-02, 3.4815e-01, 6.6317e-01,
3.3460e-01, 7.1318e-01, 1.1366e+00, 1.1671e+00, 5.2543e-02,
-8.5805e-01, 5.0632e-01, -1.0799e+00, -3.6625e-01, 6.3304e-01,
1.7650e-01, 1.2427e+00, -4.2824e-01, -9.7264e-01, -6.5294e-01,
-4.1995e-01, 1.6176e-01, 6.5306e-01, 6.5527e-01, 1.5359e-01,
-1.4403e-02, 5.5345e-01, -1.2129e+00, -1.6561e-01, -1.8614e-01,
-6.3296e-01, 9.8403e-01, 1.9044e-01, -7.3609e-01, 1.3295e-01,
-2.9614e-01, -2.6278e-01, 6.1773e-01, -2.6080e-02, -2.6567e-01,
2.1076e-01, -6.1336e-01, -3.0605e-01, 4.8003e-01, -3.7147e-01,
-3.3662e-01, 6.8647e-01, 1.2991e+00, 1.0152e+00, -4.5890e-01,
-4.8116e-01, 6.6182e-01, 2.0629e-01, 1.1687e+00, -1.4938e-01,
5.4687e-01, 2.8266e-01, 1.2739e+00, 2.1758e-01, 3.4379e-01,
-1.7554e-01, -2.9683e-01, -3.6898e-01, -3.4443e-01, 4.4539e-01,
6.2362e-01, -7.0732e-01, 6.8179e-01, -8.1357e-01, -2.9273e-02,
1.0692e-01, 3.0787e-01, -1.5126e-01, -6.9601e-01, -6.7925e-03,
-1.2032e-01, 4.3556e-01, -2.5765e-01, 3.5000e-01, 2.6138e-01,
-7.1060e-01, -7.8778e-01, -5.7865e-01, 1.9608e-02, -2.6077e-01,
2.3804e-01, 5.9406e-01, 6.5240e-01, 5.8997e-01, -2.6604e-01,
-8.2560e-01, -4.9733e-01, -1.1837e+00, -6.0205e-01, 4.3423e-01,
8.6452e-01, -6.5541e-01, -4.8626e-01, 6.9140e-01, 1.4461e-01,
-2.2926e-01, -2.0209e-01, -2.8848e-01, -9.6731e-01, -1.7899e-01,
-9.3281e-01, -1.1425e-01, 8.9797e-01, -1.6610e-01, -1.0455e+00,
-5.4111e-01, -6.5867e-01, 4.8375e-01, 8.1165e-01, -4.3111e-01,
1.2101e+00, -7.1869e-01, -9.3713e-01, 8.0531e-01, -1.0737e+00,
-8.2951e-02, 9.9555e-01, 5.7983e-01, 6.4511e-02, -1.5688e-01,
-7.8642e-01, 2.1101e-01, 3.5995e-01, -1.1792e+00, 2.7492e-01,
-1.1338e+00, 1.4390e+00, 7.4970e-01, 4.1525e-01, -2.2618e-01,
-5.6868e-01, -1.0587e+00, -8.4871e-01, 4.2294e-01, -7.5937e-01,
1.2478e+00, 3.1876e-01, 5.4925e-02, -4.2920e-01, 2.9908e-02,
-7.2298e-01, -5.0745e-01, -1.0847e-01, -3.3263e-01, 3.8415e-01,
-3.0520e-01, -7.3637e-01, 3.7017e-01, 2.0959e-01, -3.9341e-01,
9.2597e-02, -7.0634e-01, -5.4246e-01, -5.9055e-01, 1.0688e-01,
-5.5952e-01, 1.4558e+00, -7.3014e-01, -4.1277e-01, 1.1603e+00,
-2.9320e-01, -1.1457e+00, 7.0188e-01, -7.2187e-01, 4.9911e-01,
-1.7366e-01, 7.8894e-01, -6.5754e-01, -6.5171e-01, 2.3485e-01,
2.2758e-01, 1.1867e+00, 4.1541e-01, 8.7716e-02, 5.4310e-01,
-3.5418e-01, -2.2289e-01, 5.3745e-01, 4.9035e-01, -1.5985e-01,
1.4329e-02, 1.1308e+00, -9.6263e-01, 2.2490e-01, 1.6313e-01,
5.1836e-01, 2.7269e-01, 1.3424e-01, -3.2040e-01, 1.7400e-01,
-6.9929e-01, -7.7752e-01, -6.4446e-01, 2.9672e-01, -1.3011e-02,
-4.6553e-01, 5.0777e-01, -2.2849e-01, -3.6042e-01, 1.3338e+00,
1.6765e-01, -9.3721e-01, 5.1379e-01, 2.6106e-01, 7.9202e-01,
7.5039e-01, -7.1235e-02, -5.7395e-03, -5.5282e-01, 7.3725e-01,
5.6211e-01, -6.3226e-01, -4.4917e-02, 4.0115e-01, 3.1123e-01,
-8.5666e-01, -1.1569e+00, 3.8246e-01, -4.3587e-01, -1.1493e+00,
-3.8287e-01, 3.4385e-01, -7.5745e-01, -2.5882e-01, -2.8164e-01,
-1.1965e-02, 2.1589e-01, -4.3658e-01, -1.3746e-01, 9.7378e-01,
-1.0517e+00, -4.0558e-01, -1.0544e-02, -6.4660e-02, -5.1592e-01,
4.5318e-01, 1.3184e+00, 3.7338e-01, 3.4490e-02, 1.4038e+00,
-1.6802e-01, 2.3007e-02, 4.4980e-03, 1.2705e-01, 2.7906e-01,
-8.4902e-01, -3.8745e-01, 3.1278e-01, -8.8074e-01, -4.7914e-01,
7.2190e-02, -6.4725e-01, -2.0902e-01, 1.2280e-01, -1.8186e-01,
1.0589e-01, -6.0947e-01, -2.8543e-01, -1.0723e+00, -1.7837e-01,
1.4746e+00, 1.1301e+00, 8.8037e-01, -3.8367e-01, 6.1571e-01,
5.8543e-02, 5.3181e-01, -2.4058e-01, 6.9641e-01, 5.4891e-01,
4.8759e-03, -1.2818e+00, 7.0707e-01, 4.6681e-01, -2.3600e-01,
-2.7093e-01, 2.1033e-01, -3.7307e-01, 2.2353e-01, -1.5244e-01,
2.1925e-02, -2.3214e-01, 2.5308e-02, 7.7142e-01, 6.8966e-01,
5.5418e-01, -3.1878e-01, -8.5453e-01, 5.2859e-01, 7.1266e-01,
3.4018e-01, 2.4858e-01, -7.1972e-01, 1.2186e+00, -8.9309e-01,
3.7593e-01, 9.3331e-01, 1.6154e+00, 3.6179e-01, 9.8585e-01,
5.0944e-01, 3.2588e-01, -2.8218e-01, -7.9708e-02, 5.0813e-01,
7.1221e-01, 3.9624e-01, 6.6906e-01, -1.6557e-01, -4.3672e-01,
-4.2653e-01, 7.3462e-01, -8.6661e-02, 6.1583e-01, 1.1201e+00,
-1.3712e+00, 7.1885e-01, -9.1739e-02, 4.3945e-01, 1.1710e+00,
-2.2104e-01, 2.4807e-01, 2.9516e-01, -1.0306e+00, 1.2226e+00,
-2.5720e-01, -2.0021e-01, -6.3561e-01, 7.7526e-01, -1.3281e-01,
4.2105e-01, -9.8721e-01, 6.0226e-02, -7.4250e-01, -1.1144e-01,
2.1858e-01, 1.0423e+00, -2.9606e-01, 1.8390e-01, 3.6015e-01,
-3.9900e-03, 8.7758e-02, 1.1577e-01, 1.1973e+00, 3.1103e-01,
-4.2398e-01, 3.0271e-01, -2.8444e-01, 5.1662e-01, -1.3356e+00,
5.8137e-01, -5.7901e-01, 6.6809e-01, 5.4561e-01, 7.4738e-01,
-9.4391e-03, -2.7110e-01, 6.9678e-02, -3.0574e-01, 4.4471e-01,
-6.3125e-02, 4.8383e-01, -4.3376e-01, 9.3516e-01, -1.0835e+00,
3.9941e-01, 7.5706e-01, -3.8424e-02, -1.5844e-01, 6.9874e-01,
-2.2643e-01, -8.1118e-01, 3.4155e-01, 6.2197e-01, -4.5426e-01,
-1.1235e-01, -2.8171e-01, -4.2020e-01, 5.8910e-03, -2.1005e-01,
9.5020e-01, -5.1491e-01, -5.2156e-01, -3.6768e-01, -3.3160e-01,
-8.8831e-01, 1.6263e-01, -1.2790e+00, -2.2097e-01, 6.3435e-01,
3.3925e-01, 5.1287e-01, 3.0567e-02, 1.4348e-01, -3.5183e-01,
-4.7715e-01, -5.9157e-01, -5.3332e-01, -1.2875e-01, -6.4399e-01,
1.2068e+00, -4.3453e-01, 3.6951e-01, -3.7221e-01, 1.0549e+00,
-7.8913e-01, -7.2606e-01, 1.7644e+00, -7.1351e-01, -4.7304e-01,
9.7223e-01, -7.1468e-01, 1.0018e-01, 6.0553e-01, 1.4333e-01,
3.6390e-01, -5.6337e-02, 3.2207e-01, 5.4261e-01, -4.5484e-01,
-9.2550e-02, -1.0209e-01, -5.5761e-01, -8.3987e-02, 6.7479e-01,
7.3383e-01, 3.2637e-01, 2.2839e-01, 7.3619e-01, 4.5373e-02,
1.4767e+00, 1.1286e+00, 1.0320e-01, 6.0987e-03, 1.8241e-02,
-4.1522e-01, 1.4877e+00, -9.9928e-02, 1.1028e+00, 1.8680e-01,
9.3361e-01, 1.1641e-03, -6.7221e-01, -1.1105e+00, -4.5087e-01,
-1.9451e-01, 5.3225e-01, -7.0291e-02, 2.6069e-01, 3.2638e-01,
-9.7803e-01, 2.5177e-01, -5.2165e-02, 3.2999e-01, 7.0848e-01,
-4.7834e-01, 1.2501e+00, -3.3023e-01, -9.7759e-01, 6.8180e-01,
-6.7149e-01, 1.3792e+00, -3.1857e-01, -9.0531e-01, 4.8713e-01,
-1.1678e-01, -5.3198e-01, -8.2755e-01, -6.2357e-01, -7.8093e-01,
-1.9248e-01, 2.3543e-01, -3.6204e-01, -3.1342e-01, -1.2192e+00,
5.2952e-01, 2.6511e-01, 1.4131e+00, -9.5214e-02, -7.0332e-01,
-1.9167e-01, 8.7300e-01, -3.5202e-01, 8.1680e-01, -1.1795e+00,
-4.6051e-01, -7.1857e-01, -1.1671e-01, -3.5521e-01, 3.9610e-01,
6.4604e-01, 2.4028e-01, 2.4551e-01, 3.8291e-01, 3.6420e-01,
-6.4876e-01, -1.0353e-01, -3.0709e-01, 5.1353e-01, 5.7663e-01,
2.8187e-01, 3.8347e-01, -3.2358e-01, 2.4605e-01, -5.6711e-01,
7.2537e-01, 9.1548e-01, 8.8802e-01, 6.2908e-01, 3.6824e-01,
-1.1807e-01, -6.2132e-01, -2.2127e-01, 7.0661e-01, -3.4173e-01,
3.7389e-02, 6.3347e-02, -1.5815e-01, -5.4016e-02, 6.1224e-01,
-1.1764e-01, -6.6400e-01, -6.9328e-01, 1.1396e-01, 2.9845e-01,
1.1607e+00, -3.9230e-01, 6.9676e-01, 3.7753e-01, -4.5834e-01,
3.5393e-01, -1.9701e-01, -4.0007e-02, 1.6460e+00, 1.0020e+00,
2.7125e-01, 5.4337e-01, 9.6963e-01, 3.7202e-01, 1.3688e-01,
5.5945e-01, -6.7389e-01, 4.1863e-01, 9.1551e-01, -1.4212e-01,
-8.4382e-01, 6.4143e-01, -2.8692e-01, -8.3050e-01, 5.6636e-01,
3.8771e-01, 6.7161e-01, 7.1014e-01, 8.8337e-01, -6.5802e-01,
-1.9257e-01, -8.7510e-01, -8.5440e-01, -1.0751e+00, -3.9010e-01,
9.7424e-01, -1.0402e-01, 5.7751e-01, 2.9744e-01, 7.4402e-01,
1.5031e-01, 2.2013e-01, -2.3053e-01, -4.6259e-01, 2.5106e-01,
-1.0140e+00, 8.0855e-01, -1.1636e-01, -7.8362e-02, -9.2715e-02,
-1.0271e+00, 2.9693e-01, -2.0904e-01, -5.0984e-01, 1.3045e+00,
4.8532e-01, 1.4346e-01, 7.6788e-01, 9.7047e-01, 4.0762e-01,
2.3484e-01, 3.1950e-01, -2.3318e-01, -6.8306e-03, 5.6380e-01,
-9.6460e-02, -2.8250e-01, 1.1092e+00, -3.1063e-02, -2.1305e-01,
6.3479e-01, 2.1984e-01, -1.1693e+00, 4.7175e-01, -1.3506e-01,
1.1924e-01, 4.1394e-01, -1.2817e+00, -2.7704e-01, 1.0168e+00,
2.1124e-01, 6.1006e-02, -2.0014e-01, 1.4460e+00, 3.5466e-01,
3.7454e-01, -1.2640e-02, 6.0403e-03, 3.2332e-01, 8.9131e-01,
4.5607e-02, -6.6399e-02, -2.0708e+00, 3.2648e-01, 7.6369e-01,
4.1520e-01, -2.7174e-01, -5.1358e-01, 9.6802e-01, 3.8855e-01,
6.7598e-01, 3.1721e-01, 2.0969e-01, -1.3217e-01, 7.5170e-01,
-1.0165e+00, 4.6450e-01, 3.1623e-01, -1.2664e-01, -5.8193e-01,
7.5702e-01, 2.1583e-01, 8.0843e-01, 8.0445e-01, -6.3687e-01,
-2.1509e-01, 1.3130e-01, -4.3707e-01, -3.1932e-01, 2.4451e-01,
1.8980e-01, 1.7880e-01, 5.7971e-01, -9.6651e-01, 4.5083e-01,
5.5928e-01, 3.5459e-01, 1.1491e-01, 1.0462e+00, -1.2330e-01,
-2.5296e-01, 2.2241e-02, 1.1558e+00, -5.2790e-01, -5.4470e-01,
6.7174e-01, -6.3254e-01, 1.0079e-01, -5.7307e-01, -5.3185e-01,
1.3242e+00, -4.9839e-01, -5.1236e-03, -1.1210e+00, 1.0798e+00,
7.2417e-02, -4.2987e-01, 1.2662e+00, 6.0314e-01, -9.2942e-02,
9.2841e-01, -1.3543e-01, -3.3278e-01, 6.6304e-01, -4.1726e-01,
-3.8710e-02, 6.5126e-01, -7.1032e-02, 3.6333e-01, 8.8658e-02,
5.5386e-01, -1.1523e+00, -4.8741e-01, -3.0395e-01, -7.6689e-01,
1.0167e+00, 2.8526e-01, 4.7810e-01, 1.0492e-01, 3.2575e-01,
-8.6603e-01, 8.0494e-01, 6.8050e-01, 1.5820e-01, -1.0125e-01,
-9.4131e-02, -8.0925e-01, 9.6557e-01, -3.8323e-04, -5.9713e-01,
-4.7461e-01, 2.8437e-01, -7.0949e-02, -6.6371e-01, 5.8345e-01,
-1.9877e-01, -1.0992e+00, -6.4899e-01, 7.9953e-01, 4.7137e-01,
1.0099e+00, 6.3704e-01, 2.3527e-01, 2.2146e-01, -1.3238e-01,
-4.7322e-01, 1.1008e+00, 4.1789e-01, -2.6206e-01, -2.6280e-01,
-1.3215e-01, -7.2749e-01, 8.8819e-02, 8.2486e-01, 9.9206e-01]],
grad_fn=<AddmmBackward>)
tensor([[ 0.8583, 0.2219, 0.0908, -0.3688, 1.1560, 0.7270, -0.1212, -0.3032,
-0.1714, -0.0670, -1.0323, 0.0047, -1.0780, -0.8921, -1.3603, -0.1051,
-0.7071, 0.1529, 0.4977, 0.6891, -0.3599, 0.1205, -0.0934, 0.2150,
1.1170, -0.3915, 0.2290, -0.2707, 0.2720, -0.8762, -0.7861, 0.0707,
-0.3628, -0.3093, -0.5939, -0.2183, -0.0052, 0.8033, -1.1063, -0.6420,
0.1120, 0.3753, -0.4286, -0.6054, -0.1547, -0.4218, 0.3286, -0.2107,
-0.4165, -0.0471, 0.0936, -0.8109, -0.2143, -1.0776, 0.7402, -0.2014,
0.5503, -1.1897, 0.1982, 0.3422, -0.2176, -0.3140, -1.1125, 0.5685,
-0.2621, -0.0292, -0.0085, 0.8044, -0.2474, 0.0391, -0.3945, -0.3764,
1.2721, 0.0749, 0.2646, 0.0626, 0.2451, -1.3575, 0.6655, 0.0903,
-0.1422, 0.3071, 0.9701, 0.4616, 0.8723, -1.8646, 0.0220, 1.1814,
-0.9941, 1.6733, -0.0821, -0.4536, -0.0352, -0.6677, -0.1204, 1.2320,
1.1207, 0.4285, -0.4761, 0.3149, 0.2277, -0.1647, -0.4141, 0.3424,
-1.0660, 0.0641, 0.0542, -0.3486, 0.3432, 0.0687, -0.1604, -0.5909,
-1.1305, 0.3894, -0.1907, -0.7874, -0.5636, -0.4908, 0.3239, 0.3048,
0.6679, 0.7988, -0.8538, -0.2247, 0.3454, -0.9895, 0.9839, -0.4592,
-0.3357, -1.3587, 0.7747, 0.2871, -0.3907, 0.2034, 0.1411, -0.5172,
1.5404, 0.5140, -1.2516, -0.1875, -0.1076, -1.9001, 0.2219, -0.1132,
0.2477, -0.1683, 0.9590, -0.3854, -0.2032, -0.3209, -0.6780, 0.0069,
0.8684, -0.6738, -0.3045, 0.7861, 0.1697, 0.3226, -0.7716, -0.0477,
0.0432, -0.5059, 0.2730, -0.6356, 1.0412, -0.2633, -0.9977, -0.3650,
0.3670, -0.6272, -0.1175, 1.1047, 0.7568, -0.0852, -0.3786, 1.2422,
-0.5950, -0.2451, 0.5207, -0.0848, 0.0894, -0.0162, -0.4142, -0.0570,
0.1677, 0.4498, -0.8081, -0.1912, 0.4066, -0.1794, -0.1089, 0.7038,
-1.1204, 1.4453, -0.5261, -0.2537, 0.2954, 0.0537, 1.1032, -1.6995,
-0.8658, 0.0069, -0.7864, -0.5945, 0.3836, -0.8819, -0.7932, 0.7809,
-0.5035, 0.1249, -0.7372, -0.0023, 0.1401, 1.6874, -0.5839, 0.0617,
-0.3735, 0.1255, 0.8500, 0.1772, -0.2503, -0.9388, 0.4660, 0.0778,
-0.2575, 0.9906, -0.6868, -0.1666, 1.5679, -0.1536, -0.6431, 0.2470,
-0.6598, -0.3674, 1.2074, -0.5786, -1.2000, 0.5436, -0.9324, -0.1678,
0.2622, 0.2365, -1.1233, -0.0316, 0.4280, -0.6036, -0.1521, 0.8521,
-1.2506, 0.0447, -0.2429, -0.5794, 0.2477, -0.2386, -0.4713, -0.8464,
-0.6100, 0.0416, -0.9101, 0.4154, -1.1316, -0.3032, 0.2720, 0.0818,
-0.1726, 0.6396, 0.2227, -0.6746, 1.4707, 0.2891, 0.4319, 1.3665,
-1.0922, 0.2068, 0.2742, 0.5250, 0.6502, -0.8084, -0.5297, -0.3780,
-0.3048, 0.3210, -0.4358, 0.7772, -0.4798, -0.2714, -0.4301, -0.1023,
-0.8924, -0.4756, -0.9159, -0.6420, 0.2500, 0.6301, -0.3656, -0.3115,
-0.3092, -0.6765, 0.2568, -0.1190, -0.3246, -0.7433, 1.3411, 0.2621,
0.6059, -1.6666, 0.5171, 0.9830, 0.4238, 0.9399, -0.2219, 0.1042,
0.6885, -0.1398, 1.0048, 0.8237, 0.5311, 0.2481, -1.1185, -0.3169,
-0.4606, 0.2594, 0.5915, -0.0420, 0.0353, 0.5132, -0.1115, 0.1641,
0.6328, 0.0220, -0.1134, 0.3487, 0.7037, 0.1949, 0.9965, -0.3493,
-0.1531, 0.3266, 0.3365, -0.9098, 0.6647, -0.4036, 0.6309, 0.2539,
-0.6655, 0.2219, 1.3017, -0.2910, 0.0077, -0.1615, -0.0499, -1.0351,
0.0631, 0.0526, -0.2282, 0.7903, -0.2692, -0.8018, 0.0425, -0.6413,
-0.7523, -1.5698, -0.5256, -0.3538, 1.2565, 1.2090, 0.0132, -0.8184,
1.1792, -0.4623, -0.1368, -0.1340, -0.4158, -0.3891, -0.6368, -0.6716,
0.1764, -0.6001, 0.3692, 0.1826, 0.3553, -0.4659, -0.0166, -0.2227,
0.0605, -0.1283, 0.4476, -0.2427, 0.7576, 0.8014, -0.1844, 0.4134,
-0.3707, -0.5320, -0.2180, -0.7385, 0.5511, -1.1440, 0.7495, -0.1902,
0.1369, 0.2095, -0.4616, -0.2702, -0.6023, -0.1063, -0.1010, 0.4664,
-0.4199, 0.6815, 0.1581, 0.1553, -0.3236, -0.3660, -0.0891, -0.0942,
-0.8452, -1.1930, -0.7743, -0.8862, -0.5736, -0.9316, -0.1222, -0.4710,
-0.4420, -0.5289, -0.7370, -0.4200, 0.8102, 0.1068, 0.6879, 0.9414,
-1.0126, -0.6519, 0.4527, 0.3266, 0.4081, 0.1996, -0.0257, 0.1841,
-0.4881, -0.8573, 0.5010, -0.4788, -0.6908, 0.3824, -0.2642, -0.6462,
0.2921, 0.8192, 0.6443, 0.5318, 0.8571, 0.5193, -0.0748, -0.0666,
0.2659, 0.2960, -0.2691, -1.3030, -0.7433, 0.2877, -0.6012, -0.6165,
-0.1664, -0.4276, 0.7057, -0.4753, 0.9193, -0.0858, 0.4529, 0.0187,
0.5288, -0.0120, -0.2770, 0.4051, 0.0486, -0.2863, 0.7978, -0.1046,
0.5071, -0.2378, 0.0393, -0.4039, 1.1442, -0.6032, 0.6462, 0.2437,
0.6592, -0.1853, -0.3932, 0.1069, -0.3172, 0.0439, -0.0894, -0.7581,
0.8185, -0.6686, -0.7607, -0.0244, 0.0612, -0.3434, 0.4846, 0.3707,
-0.1968, 0.7238, 0.0380, -0.1852, -0.0509, 0.0693, -0.2527, -0.7352,
-0.6229, 0.4219, -0.8397, 0.0265, 0.6799, 0.2732, 0.8133, 0.5658,
0.4521, 0.2094, -0.1233, 0.2853, -0.4095, -0.0043, 0.0443, -0.1329,
0.1748, -0.3173, 0.4919, -0.2481, -0.4752, -0.3148, -0.4381, -0.8508,
0.4462, 0.6670, 0.3655, 2.1904, 0.3760, -0.1575, -0.4121, -0.8432,
0.2034, 0.7576, 0.4390, 0.1646, 0.1873, -0.3555, 0.8141, -1.1400,
-0.9080, 0.8897, 1.0261, -0.0397, 0.0434, -0.5660, 0.4741, 0.0463,
0.6119, -0.7441, -0.1215, -0.3012, 0.5099, -1.2163, -0.3103, -0.4813,
-0.3444, 0.2921, -0.9768, 0.4538, 0.6191, 0.8799, -0.3835, -0.0057,
0.2141, 0.0442, -0.3738, 0.4000, -0.9016, 0.0222, 1.1992, 0.7037,
-0.2133, 0.1393, 1.4536, -0.3438, 0.6417, 0.1219, 0.4277, -0.1041,
0.1900, 0.0222, -0.2329, -0.0655, 0.3298, -0.8072, 0.1152, -0.0886,
-0.0550, -0.1536, -0.0492, -0.2500, -0.2076, -0.3855, 0.8968, 0.2879,
0.5730, 0.1542, -0.6952, 0.6044, -0.0396, 0.6409, 1.0697, -0.5936,
-1.0671, -0.1631, -0.0559, -0.6267, -0.1045, 0.4992, -1.1814, -0.3745,
0.1148, 1.2093, 0.6348, -0.0950, 0.6317, 0.3497, 0.9094, 0.2639,
0.2698, -0.5232, 0.2271, -0.1841, -1.1478, 0.1940, 0.1754, 0.5913,
-0.1162, -0.2418, 0.2757, 0.5607, -0.3401, -0.2242, -0.5553, 0.7191,
1.1865, 0.4946, -0.0032, -0.3131, 0.4494, 0.2746, -0.0319, -0.8218,
-0.1342, 0.2442, -0.5747, -0.1053, 0.5896, -0.8873, -0.6665, 0.5551,
0.0782, 0.3987, 0.3041, 0.6591, -0.1150, -1.2871, 0.3905, 0.1369,
-0.7377, 0.9123, -0.2117, 0.4595, 0.6514, 0.4681, -0.2784, 0.0099,
-0.4514, -0.3678, -0.2100, 0.5424, 0.7370, -0.5189, 0.2916, 0.0367,
0.7997, 0.2482, -1.2903, 0.3780, 0.6159, -0.1243, -0.6554, 0.9334,
0.2720, -0.3412, 0.5984, -0.8122, -0.6410, 0.4535, 0.1734, -0.3975,
-0.8048, 0.4323, 0.4416, -0.0587, 0.3416, 0.5949, 0.9841, 0.6708,
-0.4209, -0.2902, 0.4471, -0.6019, -0.3284, 0.7052, -0.3894, 0.2325,
-0.1371, -0.0458, 0.2366, 0.6565, -0.6877, -0.4468, -0.0416, 0.1399,
0.3912, -0.5745, -0.5798, 0.3441, 0.4783, 0.6710, 1.5530, -0.2175,
0.2798, 0.7343, 0.2631, -0.1522, -0.0929, -0.7242, -0.1866, 0.4094,
0.9072, 0.7748, -0.9727, 0.2121, 0.6975, 0.5502, -1.5739, -0.1935,
0.2408, -0.9197, 0.8733, 0.1751, -1.6064, -0.8624, 0.3407, 0.1941,
0.2186, 1.0303, 0.9977, 0.9978, 0.5819, 0.3241, -0.0397, -0.0729,
-0.2124, 0.6568, 0.3392, 0.5155, -0.0025, -1.6329, 0.0523, 0.0961,
-1.1520, 0.9825, -0.1009, 0.3857, -0.6765, 0.2406, 0.7285, -0.1881,
0.3678, 0.4719, -0.5791, 0.2218, 0.6020, 0.3131, 0.2334, -0.5597,
0.7021, 0.0916, -0.0537, -1.1107, -1.3456, 0.1169, 0.2511, 0.0659,
0.3046, -0.7241, -0.0933, 0.8756, -1.0751, 0.0476, 0.7796, 0.3287,
0.5448, -0.9035, 0.5777, -0.7859, 0.4341, 1.1255, 0.4447, 0.2046,
0.6836, 0.7843, 0.9051, 0.4812, 0.2248, -0.3002, -0.2056, 0.0062,
0.6319, -0.6192, 0.1192, 0.6877, -0.1505, -0.5901, 1.1053, -0.4244,
1.2949, -0.0225, -0.7752, 0.4802, -0.8232, 0.6852, -0.3668, 0.5570,
0.6362, 0.1484, -0.1976, 0.1145, 0.4329, -0.3728, -0.0098, -0.1347,
-0.1296, -0.8788, -0.9397, -0.9967, 0.3410, -0.5902, 0.2272, 0.4113,
0.5386, -0.0544, -0.4600, -0.4375, -0.4247, -0.0638, -0.3998, -0.8774,
-0.0317, 0.4021, -0.8222, -0.4809, -1.1616, 0.8198, -0.0503, 0.5451,
-0.4983, 0.1550, -0.8350, -0.2284, 0.8163, 0.3057, -0.1393, -0.8876,
-0.1237, 0.0322, 0.0652, 0.0870, -1.2977, -0.2600, 0.3764, 0.2252,
-0.4700, 0.8265, -0.3017, -0.7971, -0.6706, 0.1718, 0.0871, -1.4964,
-0.5555, 0.2503, -0.5699, -0.6307, -0.1777, 0.0038, -1.1909, -0.9936,
0.8798, 0.3346, -0.5728, -1.2314, -0.4099, 0.3378, 0.7328, -0.1436,
-0.1719, 0.0340, -1.2943, -0.0104, 1.0345, 0.2331, 0.0599, 0.6879,
-0.7434, -0.5223, 0.2352, -0.9593, 0.0825, -0.5867, 0.5346, 0.7509,
0.8674, -0.9182, 0.3365, 0.1850, -1.1668, -1.2389, -0.3753, -0.1050,
-0.0569, 0.1062, 0.7333, 0.0465, 1.0837, 0.6633, -1.7137, -0.5974,
0.4209, -0.3124, 0.3935, 0.5566, -0.0534, -0.1313, -0.4815, 0.7935,
0.3970, -1.2496, -0.3091, -1.0784, 0.4957, 0.4002, 0.3540, -1.6088,
-0.6770, 0.7959, -0.7764, 0.7647, 0.2016, 0.3960, 0.1153, 0.3679,
0.4652, -0.1590, -0.3743, -0.0253, -0.1515, -0.7036, -0.8022, -0.2377,
-0.4536, 0.2447, 0.2165, 0.0511, -1.0900, -0.3818, -0.9283, 0.4730,
0.4143, 0.0216, 0.1163, -0.1247, 0.2278, -0.6479, -0.5509, 0.5441,
0.2503, -0.0678, 0.8512, 0.3365, -0.5701, 0.1218, 0.2744, -0.7122]],
grad_fn=<AddmmBackward>) Process finished with exit code
陈云pytorch学习笔记_用50行代码搭建ResNet的更多相关文章
- 50行代码实现GAN | 干货演练
2014年,Ian Goodfellow和他的同事发表了一篇论文,向世界介绍了生成对抗网络(GAN).通过对计算图和博弈论的创新性组合,他们表明如果有足够的建模能力,两个相互对抗的模型可以通过普通的反 ...
- python3.4学习笔记(五) IDLE显示行号问题，插件安装和其他开发工具介绍
python3.4学习笔记(五) IDLE显示行号问题,插件安装和其他开发工具介绍 IDLE默认不能显示行号,使用ALT+G 跳到对应行号,在右下角有显示光标所在行.列.pycharm免费社区版.Su ...
- jQuery源代码学习笔记_工具函数_noop/error/now/trim
jQuery源代码学习笔记_工具函数_noop/error/now/trim jquery提供了一系列的工具函数,用于支持其运行,今天主要分析noop/error/now/trim这4个函数: 1.n ...
- 【学习笔记】OI玄学道—代码坑点
[学习笔记]\(OI\) 玄学道-代码坑点 [目录] [逻辑运算符的短路运算] [\(cmath\)里的贝塞尔函数] 一:[逻辑运算符的短路运算] [运算规则] && 和 || 属于逻 ...
- python3.4学习笔记(十三) 网络爬虫实例代码，使用pyspider抓取多牛投资吧里面的文章信息，抓取政府网新闻内容
python3.4学习笔记(十三) 网络爬虫实例代码,使用pyspider抓取多牛投资吧里面的文章信息PySpider:一个国人编写的强大的网络爬虫系统并带有强大的WebUI,采用Python语言编写 ...
- python爬虫实战：利用scrapy，短短50行代码下载整站短视频
近日,有朋友向我求助一件小事儿,他在一个短视频app上看到一个好玩儿的段子,想下载下来,可死活找不到下载的方法.这忙我得帮,少不得就抓包分析了一下这个app,找到了视频的下载链接,帮他解决了这个小问题 ...
- JPG学习笔记3（附完整代码）
#topics h2 { background: rgba(43, 102, 149, 1); border-radius: 6px; box-shadow: 0 0 1px rgba(95, 90, ...
- Java web与web gis学习笔记（一）——Tomcat环境搭建
系列链接: Java web与web gis学习笔记(一)--Tomcat环境搭建 Java web与web gis学习笔记(二)--百度地图API调用 JavaWeb和WebGIS学习笔记(三)-- ...
- openstack学习笔记一 虚拟机启动过程代码跟踪
openstack学习笔记一 虚拟机启动过程代码跟踪 本文主要通过对虚拟机创建过程的代码跟踪.观察虚拟机启动任务状态的变化,来透彻理解openstack各组件之间的作用过程. 当从horizon界面发 ...
随机推荐
- js中数据操作的某些思想
1,默认数据的复用 写成function的return形式缓存在变量中,用的时候直接执行fun就行了 例如 有文件text.js里面的对象是export default ()=>({aa:55, ...
- shell编写启动脚本
[root@confluence bin]# vim /etc/init.d/confluence #!/bin/bash # Confluence Linux service controller ...
- 08-01-json-loggin-模块
复习 ''' ATM: -- start.py BASE_DIR = os.path.dirname(__file__) sys.path.append(BASE_DIR) -- conf -- li ...
- SpringBoot - @ControllerAdvice 处理异常
在Spring 3.2中,新增了@ControllerAdvice.@RestControllerAdvice 注解,可以用于定义@ExceptionHandler.@InitBinder.@Mode ...
- javascript的简洁的写法
每种语言都有它特别的地方,对于JavaScript来说,使用var就可以声明任意类型的变量,这门脚本语言看起来很简单,然而想要写出优雅的代码却是需要不断积累经验的.本文利列举了JavaScript初学 ...
- 【LeetCode 12】整数转罗马数字
题目链接 [题解] (涨知识了..原来罗马数字是这么回事.. 把{1,4,5,9,10,40,50,90,100,400,500,900,1000}这些东西的罗马数字放在一个数组里面. 每次从大到小减 ...
- C/S模式简单socket通信
TCP连接方式 sever.c #include <stdio.h>#include <stdlib.h>#include <sys/socket.h>#inclu ...
- CSS：CSS 边框
ylbtech-CSS:CSS 边框 1.返回顶部 1. CSS 边框 CSS 边框属性 边框样式 边框样式属性指定要显示什么样的边界. border-style属性用来定义边框的样式 border ...
- Guava EventBus集成spring
EventBus 不是通用的消息系统,也不是用来做进程间的通信的,而是在进程内,用于解耦两段直接调用的业务逻辑: 1.代码结构 event:eventbus中流转的事件(消息),包结构按照业务模块在细 ...
- 文本表征：SoW、BoW、TF-IDF、Hash Trick、doc2vec、DBoW、DM
原文地址:https://www.jianshu.com/p/2f2d5d5e03f8 一.文本特征 (一)基本文本特征提取 词语数量 常,负面情绪评论含有的词语数量比正面情绪评论更多. 字符数量 常 ...