您现在的位置是:貌合心离网 > 热点
Pytorch中的NN模块并实现第一个神经网络模型
貌合心离网2026-11-14【热点】{固定数字1}人已围观
简介谷歌霸屏外推技术【tg:seogoogle987】
在PyTorch建立模型,主要是NN模块。
nn.Linear
nn.Linear是创建一个线性层。这里需要将输入和输出维度作为参数传递。
linear = nn.Linear(10, 2) example_input = torch.randn(3, 10) example_output = linear(example_input) example_output
上面代码linear接受nx10的输入并返回nx2的输出。
print(example_input) print(example_output) tensor([[ 1.1122, -0.1381, 0.5547, -0.3326, Py-0.5676, 0.2810, -0.5521, -0.8729, -0.6627, 0.8729], [ 1.9134, 0.2397, -0.8340, 1.1532, -1.6725, 0.6171, -0.0357, -1.6848, -0.8454, 0.3876], &ntobsp; [-0.0786, -0.1541, -0.8385, -0.1587, -0.0121, 1.4457, -0.0132, 1.5653, -1.6954, -0.9350]]) # 输出如下 tensor([[-0.1249, -0.8002], [-1.0945, -0.2297], &rcnbsp; [-0.3558, 0.8439]], grad_fn=) nn.Relu
nn.Relu对线性的给定输出执行 relu 激活函数操作。
relu = nn.ReLU() relu_output = relu(example_output) relu_output # 输出如下 tensor([[0.0000, 0.0000], [0.0000, 0.0000], [0.0000, 0.h中8439]], grad_fn=) nn.BatchNorm1d
nn.BatchNorm1d是一种标准化技术,用于在不同批次的输入中保持一致的均值和标准偏差。
batchnorm = nn.BatchNorm1d(2) batchnorm_output = batchnorm(relu_output) batchnorm_output # 输出如下 tensor([[ 0.0000, -0.7071], [ 0.0000, -0.7071], &nbs的Np; [ 0.0000, 1.4142]], grad_fn=) nn.Sequential
nn.Sequential一次性创建一系列操作。和tensorflow中的Sequential完全一样。
mlp_layer = nn.Sequential( nn.Linear(5, 2), nn.BatchNorm1d(2), nn.ReLU() ) test_example = torch.randn(5,5) + 1 print("input: ") print(test_N模example) print("output: ") print(mlp_layer(test_example)) # 输出如下 input: tensor([[ 1.4617, 1.2446, 1.4919, 1.5978, -0.3410], [-0.2819, 0.5567, 1.0113, 1.8053, -0.0833], [ 0.2830,&nb块并sp; 1.0857, 1.2258, 2.6602, 0.1339], [ 0.8682, 0.9344, 1.3715, 0.0279, 1.8011], [ 0.6172, 1.1414, 0.6030, 0.3876, 1.3653]]) output: tensor([[0.0000, 0.0000实现], [0.0000, 1.3722], [0.0000, 0.8861], [1.0895, 0.0000], [1.3047, 0.0000]], grad_fn=)
在上面的模型中缺少了优化器,我们无法得到对应损失。
import torch.optim as optim adam_opt = 第一;optim.Adam(mlp_layer.parameters(), lr=1e-1) # 这里lr表示学习率,1e-1表示0.1 train_example = torch.randn(100,5) + 1 adam_opt.zero_grad() # 我们将使用1减去平均值,作为简单损失函数 cur_loss = torch.abs(1 - mlp_layer(train_example)).mean() cur_loss.backward() # 更新参数 adam_opt.step() print(cur_loss.data) # 输出如下 tensor(0.7467)
虽然个神上面只是用了一个epoch,训练线性模型得到loss为0.7467,上面就是NN模型建立model的整个流程,
第一个神经网络模型
下面实现第一个分类神经网络,其中一个隐藏层用于开发单个输出单元。
首先,使用以下命令导入 PyTorch 库 -
import torch import torch.nn as nn
定义所有层和批量大小以开始执行神经网络,如下所示 -
n_in, n_h, n_out, batch_size = 10, 5, 1, 10
由于神经网络包括输入数据的组合以获得相应的输出数据,我们将遵循以下相同的程序 -
x = torch.randn(batch_size, n_in) y = torch.ten经网sor([[1.0], [0.0], [0.0], [1.0], [1.0], [1.0], [0.0], [0.0], [1.0], [1.0]])
创建顺序模型。使用下面代码,创建一个顺序模型 -
model = nn.Sequential(nn.Linear(n_in, n_h), nn.ReLU(), nn.Linear(n_h, n_out), nn.Sigmoid())
借助梯度下降优化器构建损失函数,如下所示 -
# 构造损失函数 criterion = torch.nn.MSELoss络模() # 构造优化器 optimizer = torch.optim.SGD(model.parameters(), lr = 0.01)
使用给定代码行的迭代循环实现梯度下降模型 -
# 梯度下降 for epoch in range(50): # 正向传递:通过将x传递给模型来计算预测的y y_pred = model(x) # 计算loss loss = criterion(y_pred, y)  Py; # 梯度清0 optimizer.zero_grad() # 反向传播,求解梯度 loss.backward() # 更新模型参数 optimizer.step() if epoch % 10 == 0: print('epoch: ', epoch,' loss: ', loss.item())&tonbsp;
输出如下
epoch: 0 loss: 0.2508794665336609 epoch: 10 loss: 0.24847669899463654 epoch: 20 loss: 0.24615907669067383 epoch: 30 loss: 0.24392127990722656 epoch: 40 loss: 0.24175791442394257
很赞哦!(8285)
相关文章
- 标签: 字节跳动 沐瞳 MobileLegends Savvy
- 阿里云三季度营收276.48亿元 同比增长2%
- 特斯拉回应“机场禁止入内”:哨兵模式数据仅存储在车辆本地 另外特斯拉已在中国建立数据中心
- Google Earth AI“地球大脑”上线!Gemini一键预测洪水、藻华、野火,全球灾害预警从数周缩短至几分钟
- 创新平台 Emergent 获2300万美元 A 轮融资,助力无编程经验者开发应用
- OpenAI与Anthropic的阴影:Cursor500亿美元融资为何遭大厂投资者拒绝
- 标签: 安克创新
- 工信部等六部门:到2025年新建大型和超大型数据中心PUE优于1.3
- 用户警惕!Anthropic Claude Desktop 被指悄悄安装间谍软件
- 标签: 俞敏洪 新东方文旅 东方甄选 知识文旅
热门文章
站长推荐
友情链接
- 标签: 康师傅
- https://www.kuaikanmanhua.com/sou/谷歌霸屏外推技术【tg:seogoogle987】,9gn
- https://www.teamlab.art/fr/?submit=谷歌霸屏外推技术【tg:seogoogle987】,fos
- https://www.totolink.vn/search/谷歌霸屏外推技术【tg:seogoogle987】,bYk
- https://www.sportbiketracktime.com/index.php?route=product/search&tag=谷歌霸屏外推技术【tg:seogoogle987】,1aj
- https://www.l-and-b.jp/?page=1&s=谷歌霸屏外推技术【tg:seogoogle987】,v0M
- https://www.ons.gov.uk/search?q=谷歌霸屏外推技术【tg:seogoogle987】,2UV
- https://www.temu.com/search_result.html?search_key=谷歌霸屏外推技术【tg:seogoogle987】DTE
- https://www.mexperience.com/?T=1686132112324&s=谷歌霸屏外推技术【tg:seogoogle987】,7b0
- https://www.rookwinkel.nl/catalogsearch/result/index/?q=谷歌霸屏外推技术【tg:seogoogle987】,D8A
- https://www.remit.co.jp/vi/?s=谷歌霸屏外推技术【tg:seogoogle987】13C
- https://www.qidulp.com/search?query=谷歌霸屏外推技术【tg:seogoogle987】,xKQ
- https://www.teacherspayteachers.com/Browse/Price-Range/Free/Search:Western International University Physics degree=谷歌霸屏外推技术【tg:seogoogle987】,RG6
- https://www.lifesciencesweden.se/search?q=谷歌霸屏外推技术【tg:seogoogle987】,rYg
- https://www.maafushidive.com?collections?vendors?q=谷歌霸屏外推技术【tg:seogoogle987】,pOp
- https://www.paxnet.co.kr/search/jongmok?kwd=谷歌霸屏外推技术【tg:seogoogle987】I50
- https://www.dba.dk/soeg/?soeg=谷歌霸屏外推技术【tg:seogoogle987】DcF
- https://www.miamiopenmri.com/?action=CHECK&s=谷歌霸屏外推技术【tg:seogoogle987】,7Br
- https://store.line.me/search/sticker/ja?q=谷歌霸屏外推技术【tg:seogoogle987】T7M
- https://www.ushmm.org/search/results.php?q=谷歌霸屏外推技术【tg:seogoogle987】,pmZ
- https://www.lneg.pt/en/?s=谷歌霸屏外推技术【tg:seogoogle987】,cSf
- https://www.tombow.com/en/search/谷歌霸屏外推技术【tg:seogoogle987】,DJK
- https://www.usslexington.com/?action=QUERY&s=谷歌霸屏外推技术【tg:seogoogle987】,Puz
- https://www.sbaad.no/?s=谷歌霸屏外推技术【tg:seogoogle987】,nSF
- https://www.spicebar.de/catalogsearch/result/?q=谷歌霸屏外推技术【tg:seogoogle987】,HW7
- https://www.whatsonstage.com/search?cat=64&term=谷歌霸屏外推技术【tg:seogoogle987】,Kl4
- https://www.tombow.com/search/谷歌霸屏外推技术【tg:seogoogle987】,vtS
- https://www.pinkpanda.ro/magazin/ingrijirea-tenului?page=87&q=谷歌霸屏外推技术【tg:seogoogle987】,vDa
- https://www.universal-music.co.jp/search/?q=谷歌霸屏外推技术【tg:seogoogle987】,b22
- https://blvdseafood.com/?s=谷歌霸屏外推技术【tg:seogoogle987】gtv
- https://search.danawa.com/dsearch.php?query=谷歌霸屏外推技术【tg:seogoogle987】2Wi
- https://www.msrenewal.com/search/谷歌霸屏外推技术【tg:seogoogle987】,ov7
- https://www.revere.org/search?action=SEARCH&q=谷歌霸屏外推技术【tg:seogoogle987】,Iye
- https://www.schenckprocess.com/search?q=谷歌霸屏外推技术【tg:seogoogle987】,2uq
- https://www.yidian51.com/taobao/谷歌霸屏外推技术【tg:seogoogle987】,Cu9
- https://www.mintos.org/?s=谷歌霸屏外推技术【tg:seogoogle987】,Jye
- https://www.metaltix.com/?view=tours&filter=search&q=谷歌霸屏外推技术【tg:seogoogle987】,bbd
- https://www.ng-slo.si/si/iskanje?cat=114&q=谷歌霸屏外推技术【tg:seogoogle987】,HQl
- https://www.paspaley.com/catalogsearch/result/?action=QUERY&q=谷歌霸屏外推技术【tg:seogoogle987】,uSI
- https://www.reginatangoshoes.com/en/catalogsearch/result/index/?T=1686132113407&q=谷歌霸屏外推技术【tg:seogoogle987】,zsZ
- https://yokohama-oht.com/?action=CHECK&s=谷歌霸屏外推技术【tg:seogoogle987】,ZmS
- https://www.manga-news.com/index.php/recherche/?q=谷歌霸屏外推技术【tg:seogoogle987】,zGu
- https://www.takachi-enclosure.com/search?eqYB=gnI31Ou&text=谷歌霸屏外推技术【tg:seogoogle987】,5IK
- https://www.tahiti-infos.com/search/谷歌霸屏外推技术【tg:seogoogle987】,udH
- https://www.ushmm.org/search/results/?ID=8s0587Oi8UTv74zPPM4fuhwjNpe2hsEE&q=谷歌霸屏外推技术【tg:seogoogle987】,cro
- https://www.nbm.org/?action=QUERY&s=谷歌霸屏外推技术【tg:seogoogle987】,Jhn
- https://www.zisha.com/shop/list.shtml?wap_keyword=谷歌霸屏外推技术【tg:seogoogle987】,aoP
- https://www.whirlpool.com.hk/chinese/catalogsearch/result?cat=30&q=谷歌霸屏外推技术【tg:seogoogle987】,Bvu
- https://www.swanagerailway.co.uk/search?q=谷歌霸屏外推技术【tg:seogoogle987】,UgS
- https://www.wmfe.org/?s=谷歌霸屏外推技术【tg:seogoogle987】,ijQ
- https://www.sensorshop24.de/catalogsearch/result/index/?type=132&q=谷歌霸屏外推技术【tg:seogoogle987】,oO7







