Iam_Human的博客

1. 定义与属性0维张量（标量 scalar）：0，11维张量（数组/向量 vector）：[0], [0, 1, 2]2维张量（矩阵 matrix）：3维张量 tensor…1.1 定义与访问定义# torch.tensor(data, dtype=None, device=None, requires_grad=False, pin_memory=False) → Tensora_0d = torch.tensor(0)a_1d = torch.tensor([0]) a_2d

1. 基本数据类型2. 类型转换2.1 CPU和GPU的Tensor之间转换从cpu –> gpu，使用data.cuda()即可。若从gpu –> cpu，则使用data.cpu()。2.2 Tensor与Numpy Array之间的转换Tensor –> Numpy.ndarray 可以使用 data.numpy()，其中data的类型为torch.Tensor。Numpy.ndarray –> Tensor 可以使用torch.from_numpy(data)

1. 求导# Only Tensors of floating point dtype can require gradientsx = torch.ones(1, 2, requires_grad=True)y = torch.tensor([[1.5], [2.]], requires_grad=True)# print(x.grad_fn)# r = x1*y1 + x2*y2 ...

import torchimport torch.nn as nnimport torch.nn.functional as F# Define modelclass TheModelClass(nn.Module): def __init__(self): super(TheModelClass,self).__init__() self.fc1=...

tool.py"""6.2 rnn """def test_P154(): X, W_xh = torch.randn(3, 1), torch.randn(1, 4) H, W_hh = torch.randn(3, 4), torch.randn(4, 4) R1 = torch.matmul(X, W_xh) + torch.matmul(H, W_hh) R2 = torch....

tool.py#coding=utf-8import torchimport torch.nn as nnimport torch.nn.functional as Fimport torchvisionimport torchvision.transforms as transformsimport numpy as npimport collectionsimport ...

#coding=utf-8import collectionsimport mathimport osimport randomimport sysimport tarfileimport timeimport zipfile# from tqdm import tqdm# from IPython import displayimport numpy as npfrom...

Module例1class MyNet(torch.nn.Module):def __init__(self): super(MyNet, self).__init__() self.conv1 = torch.nn.Conv2d(3, 32, 3, 1, 1) self.relu1 = torch.nn.ReLU() self.max_pooling1=torch.nn.Max...

Learning notes for Pytorch （1）torch.tensor1. constructionconstruct from list/numpydata: Can be a list, tuple, NumPy ndarray, scalar, and other typestorch.tensor([[1., -1.], [1., -1.]])>>&...