在pytorch中为Module和Tensor指定GPU的例子
pytorch指定GPU
在用pytorch写CNN的时候，发现一运行程序就卡住，然后cpu占用率100%，nvidia-smi查看显卡发现并没有使用GPU。所以考虑将模型和输入数据及标签指定到gpu上。
pytorch中的Tensor和Module可以指定gpu运行，并且可以指定在哪一块gpu上运行，方法非常简单，就是直接调用Tensor类和Module类中的.cuda()方法。
importtorch fromPILimportImage importtorch.nnasnn importnumpyasnp fromtorch.autogradimportVariable #先看看有没有显卡 torch.cuda.is_available() Out[16]:True #嗯，有显卡，可以指定，先生成一个Tensor a=torch.Tensor(3,5) a Out[13]: .00000e-05* 0.00000.00002.04190.00002.0420 0.00000.00000.00000.00000.0000 0.01320.00000.01310.00000.0000 [torch.FloatTensorofsize3x5] a.cuda() Out[14]: .00000e-05* 0.00000.00002.04190.00002.0420 0.00000.00000.00000.00000.0000 0.01320.00000.01310.00000.0000 [torch.cuda.FloatTensorofsize3x5(GPU0)] #可以看到上面显示了（GPU0）,也就是说这个Tensor是在第一个GPU上的 a.cuda(1) Traceback(mostrecentcalllast): File"",line1,in a.cuda(1) File"/home/chia/anaconda2/lib/python2.7/site-packages/torch/_utils.py",line57,in_cuda withtorch.cuda.device(device): File"/home/chia/anaconda2/lib/python2.7/site-packages/torch/cuda/__init__.py",line127,in__enter__ torch._C._cuda_setDevice(self.idx) RuntimeError:cudaruntimeerror(10):invaliddeviceordinalattorch/csrc/cuda/Module.cpp:84 #这个报错了，因为只有一块GPU，所以指定cuda(1)无效。
同样滴，Variable变量和Module类型的模型也可以指定放在哪块GPU上
v=Variable(a) v Out[18]: Variablecontaining: .00000e-05* 0.00000.00002.04190.00002.0420 0.00000.00000.00000.00000.0000 0.01320.00000.01310.00000.0000 [torch.FloatTensorofsize3x5] v.cuda(0) Out[19]: Variablecontaining: .00000e-05* 0.00000.00002.04190.00002.0420 0.00000.00000.00000.00000.0000 0.01320.00000.01310.00000.0000 [torch.cuda.FloatTensorofsize3x5(GPU0)] model=DenoiseCNN() model Out[22]: DenoiseCNN( (hid_layer):Sequential( (0):Conv2d(32,32,kernel_size=(3,3),stride=(1,1),padding=(1,1)) (1):BatchNorm2d(32,eps=1e-05,momentum=0.1,affine=True) (2):LeakyReLU(0.2) ) (first_layer):Sequential( (0):Conv2d(1,32,kernel_size=(3,3),stride=(1,1),padding=(1,1)) (1):LeakyReLU(0.2) ) (last_layer):Sequential( (0):Conv2d(32,1,kernel_size=(3,3),stride=(1,1),padding=(1,1)) ) ) model.cuda(0) Out[23]: DenoiseCNN( (hid_layer):Sequential( (0):Conv2d(32,32,kernel_size=(3,3),stride=(1,1),padding=(1,1)) (1):BatchNorm2d(32,eps=1e-05,momentum=0.1,affine=True) (2):LeakyReLU(0.2) ) (first_layer):Sequential( (0):Conv2d(1,32,kernel_size=(3,3),stride=(1,1),padding=(1,1)) (1):LeakyReLU(0.2) ) (last_layer):Sequential( (0):Conv2d(32,1,kernel_size=(3,3),stride=(1,1),padding=(1,1)) ) )
这样看不出来Module的变化，考虑看一下Module中的参数在哪里
fori,parainenumerate(model.parameters()): ifi<2: printpara Parametercontaining: (0,0,.,.)= -3.1792e-02-4.6396e-02-4.3472e-02 3.4903e-021.8558e-025.3955e-03 2.4986e-023.8061e-02-1.6658e-02 (0,1,.,.)= -3.5041e-02-3.6286e-02-3.0819e-02 1.0683e-029.0773e-03-2.5379e-02 2.9508e-032.8774e-027.4632e-04 (0,2,.,.)= -4.6986e-02-5.1183e-028.4346e-04 -6.6864e-03-2.8816e-021.2566e-02 2.1682e-022.5485e-02-7.2600e-03 ... (0,29,.,.)= -5.5289e-03-2.6012e-02-2.7771e-02 2.7528e-023.0460e-02-1.2829e-02 7.3387e-035.2633e-025.0601e-02 (0,30,.,.)= -3.5881e-029.7000e-03-3.3692e-02 1.6257e-03-4.0113e-023.5300e-02 -2.1399e-033.0934e-02-2.7513e-02 (0,31,.,.)= -2.7492e-022.5803e-025.2171e-02 -2.4082e-023.1887e-021.1292e-02 5.8893e-02-3.5452e-02-1.2115e-02 ⋮ (1,0,.,.)= 5.0664e-02-4.1085e-022.9089e-02 2.1555e-025.7176e-02-7.5013e-03 3.5075e-02-1.6610e-023.4904e-02 (1,1,.,.)= 4.6716e-02-1.2552e-02-3.8132e-02 -2.9573e-02-3.5008e-02-4.2891e-02 9.5230e-03-4.8599e-022.5357e-02 (1,2,.,.)= -1.7859e-021.3442e-021.9493e-02 1.8434e-021.4884e-038.6479e-03 -7.1610e-033.5724e-026.2249e-03 ... (1,29,.,.)= -3.3194e-021.6803e-052.3405e-02 -5.2223e-026.5680e-03-1.8427e-02 -1.4476e-02-1.5434e-02-2.3108e-02 (1,30,.,.)= 2.3479e-021.2840e-024.5949e-02 4.4833e-024.9272e-02-3.7634e-02 4.2787e-028.5841e-041.2332e-02 (1,31,.,.)= 4.1723e-02-2.5295e-021.1326e-02 -5.1707e-025.3201e-024.8928e-02 -1.6735e-02-8.7450e-03-4.9530e-02 ⋮ (2,0,.,.)= -3.1728e-02-3.9757e-026.5561e-03 -1.7731e-022.8615e-022.7457e-02 -2.1817e-03-4.2405e-02-3.6126e-03 (2,1,.,.)= 3.2434e-02-1.1574e-031.3353e-02 -2.3069e-024.9532e-021.6768e-02 -3.5563e-02-4.4264e-02-2.0571e-02 (2,2,.,.)= 7.4980e-03-5.7412e-03-3.0638e-03 1.1812e-02-1.7851e-024.2195e-04 3.9753e-023.8771e-024.3166e-03 ... (2,29,.,.)= -5.0798e-024.3651e-02-2.3798e-02 -6.0957e-03-5.6953e-021.2583e-02 -2.3450e-02-4.7136e-025.2458e-02 (2,30,.,.)= 1.5088e-022.6097e-024.9392e-03 -9.0372e-03-5.3276e-02-1.7824e-02 3.2060e-035.8820e-021.3459e-02 (2,31,.,.)= -5.2557e-03-4.9638e-02-7.5522e-03 2.8668e-02-3.9617e-02-1.8111e-02 -4.0412e-021.1320e-02-2.4005e-02 ⋮ (29,0,.,.)= -1.4393e-022.1343e-025.1940e-02 5.7449e-023.1327e-02-1.0721e-02 -1.0184e-02-6.2289e-033.9823e-02 (29,1,.,.)= -4.2240e-035.8135e-025.2816e-02 -4.9888e-023.3972e-024.3127e-02 -2.3355e-02-5.5401e-023.4952e-02 (29,2,.,.)= 4.0336e-027.6532e-035.4083e-02 -2.7456e-023.9090e-024.4008e-02 -2.0424e-02-5.8922e-02-4.4759e-03 ... (29,29,.,.)= 8.8037e-031.0347e-02-2.2285e-02 -1.0538e-02-3.2981e-022.2300e-02 -2.7337e-025.3113e-025.4608e-02 (29,30,.,.)= 3.1429e-025.2024e-03-1.3882e-02 -3.3123e-02-2.7633e-031.9007e-02 -2.9795e-023.7551e-025.6486e-02 (29,31,.,.)= 2.0140e-021.8530e-027.4208e-03 2.7311e-025.3581e-02-2.5553e-02 -1.7285e-021.8722e-024.0104e-02 ⋮ (30,0,.,.)= 5.2750e-024.5757e-03-5.3894e-02 -3.9297e-023.2918e-022.3571e-02 -1.1806e-021.6091e-023.3755e-04 (30,1,.,.)= 4.2858e-02-5.2211e-02-3.5660e-02 1.4807e-02-5.8873e-025.5535e-02 4.9854e-022.2946e-024.0968e-03 (30,2,.,.)= 3.0378e-022.1315e-029.1700e-03 3.6277e-02-4.0734e-024.8175e-02 3.0748e-02-2.7425e-02-1.7741e-02 ... (30,29,.,.)= 3.1883e-022.5012e-022.8504e-02 -1.3538e-023.5570e-02-2.0261e-02 -1.5959e-023.3373e-028.3261e-03 (30,30,.,.)= 2.7152e-02-5.6752e-022.2697e-02 1.2614e-02-2.4174e-02-2.5058e-02 1.8737e-02-1.3581e-03-3.7116e-02 (30,31,.,.)= -4.3278e-022.5873e-02-1.6677e-02 3.9483e-025.7898e-02-4.1450e-02 -5.8218e-02-3.0660e-02-4.2161e-02 ⋮ (31,0,.,.)= 1.3370e-02-1.4191e-02-2.2524e-02 2.1772e-02-2.2440e-02-3.0512e-03 3.4139e-02-1.9043e-021.1289e-02 (31,1,.,.)= -5.1293e-02-5.2802e-021.7022e-02 5.1031e-02-1.0345e-02-4.4780e-02 -4.9422e-024.7709e-02-2.1215e-02 (31,2,.,.)= 2.2289e-02-2.1746e-02-5.3192e-02 2.6651e-02-1.6531e-022.2640e-02 1.4012e-021.1405e-02-1.4809e-02 ... (31,29,.,.)= 2.5505e-032.4052e-02-4.7662e-02 1.6068e-02-4.2278e-02-2.4670e-02 -1.4684e-02-3.8222e-02-5.0006e-02 (31,30,.,.)= -4.9350e-024.7564e-02-7.3479e-03 2.6490e-02-1.1745e-023.4324e-02 4.2650e-02-5.4633e-029.4581e-03 (31,31,.,.)= -3.2695e-02-2.8899e-021.5543e-02 -5.3662e-025.0727e-023.5950e-02 4.6130e-02-4.4754e-02-4.5647e-02 [torch.cuda.FloatTensorofsize32x32x3x3(GPU0)] Parametercontaining: .00000e-02* -1.2723 -5.2970 -3.4638 -1.5302 0.7641 5.7516 -4.8427 -0.7230 4.5940 -4.1709 4.8093 -4.7249 -2.2756 -5.5165 5.1259 -2.4693 1.8527 -0.4210 -2.0518 -3.8124 -4.6195 -4.3019 -0.8631 -0.4400 5.4604 -5.5597 1.5557 4.2336 3.9482 -1.4457 2.6124 -1.8218 [torch.cuda.FloatTensorofsize32(GPU0)]
可以看出，模型的参变量是放在GPU上的。
通过指定了gpu后，就可以使用gpu来训练模型了～美滋滋
以上这篇在pytorch中为Module和Tensor指定GPU的例子就是小编分享给大家的全部内容了，希望能给大家一个参考，也希望大家多多支持毛票票。