https://blog.overfit.org/tags/pytorch/Recent content in PyTorch on 里列昂遗失的记事本Hugozh-cnFri, 15 Aug 2025 16:01:00 +0800https://blog.overfit.org/posts/2025-08-15-19039821-lenetpytorch%E5%AE%9E%E7%8E%B0/Fri, 15 Aug 2025 15:59:00 +0800https://blog.overfit.org/posts/2025-08-15-19039821-lenetpytorch%E5%AE%9E%E7%8E%B0/LeNet LeNet，它是最早发布的卷积神经网络之一，因其在计算机视觉任务中的高效性能而受到广泛关注。 这个模型是由AT&T贝尔实验室的研究员Yann LeCun在1989年提出的（并以其命名），目的是识别图像 (LeCun et al., 1998)中的手写数字。 当https://blog.overfit.org/posts/2023-01-29-17073008-pytorch%E5%B8%B8%E7%94%A8%E4%BB%A3%E7%A0%81%E5%9D%97/Sun, 29 Jan 2023 16:14:00 +0800https://blog.overfit.org/posts/2023-01-29-17073008-pytorch%E5%B8%B8%E7%94%A8%E4%BB%A3%E7%A0%81%E5%9D%97/PyTorch最好的资料是官方文档。 1. 基本配置 需要导入的包 import collections import os import shutil import tqdm import numpy as np import PIL.Image import torch imhttps://blog.overfit.org/posts/2022-05-18-16286033-%E7%8C%AB%E7%8B%97%E6%95%B0%E6%8D%AE%E9%9B%86%E5%88%86%E7%B1%BBvgg16tensorflow%E5%AE%9E%E7%8E%B0/Wed, 18 May 2022 19:33:00 +0800https://blog.overfit.org/posts/2022-05-18-16286033-%E7%8C%AB%E7%8B%97%E6%95%B0%E6%8D%AE%E9%9B%86%E5%88%86%E7%B1%BBvgg16tensorflow%E5%AE%9E%E7%8E%B0/import random import os import tensorflow as tf from tensorflow import keras from shutil import copyfile from tensorflow.keras import layershttps://blog.overfit.org/posts/2022-05-18-16285858-%E7%8C%AB%E7%8B%97%E6%95%B0%E6%8D%AE%E9%9B%86%E5%88%86%E7%B1%BBvgg16pytorch%E5%AE%9E%E7%8E%B0/Wed, 18 May 2022 18:34:00 +0800https://blog.overfit.org/posts/2022-05-18-16285858-%E7%8C%AB%E7%8B%97%E6%95%B0%E6%8D%AE%E9%9B%86%E5%88%86%E7%B1%BBvgg16pytorch%E5%AE%9E%E7%8E%B0/数据集地址：https://www.kaggle.com/datasets/shaunthesheep/microsoft-catsvsdogs-dataset ```python from shutil import copyfile import random importhttps://blog.overfit.org/posts/2021-09-06-15236192-pytorch%E7%8E%AF%E5%A2%83%E7%9A%84%E9%85%8D%E7%BD%AE/Mon, 06 Sep 2021 23:01:00 +0800https://blog.overfit.org/posts/2021-09-06-15236192-pytorch%E7%8E%AF%E5%A2%83%E7%9A%84%E9%85%8D%E7%BD%AE/PyTorch的配置 这里附上官网的链接https://pytorch.org/get-started/locally/ CPU版本 选择PyTorch版本(PyTorch Build)，Stable(稳定版)，Nightly(测试版) 系统版本，这里只介绍Windows系统的Phttps://blog.overfit.org/posts/2021-09-06-15236094-mnist%E6%89%8B%E5%86%99%E6%95%B0%E5%AD%97%E9%9B%86%E7%9A%84%E5%A4%9A%E5%88%86%E7%B1%BB%E9%97%AE%E9%A2%98convolution-neural-network/Mon, 06 Sep 2021 22:34:00 +0800https://blog.overfit.org/posts/2021-09-06-15236094-mnist%E6%89%8B%E5%86%99%E6%95%B0%E5%AD%97%E9%9B%86%E7%9A%84%E5%A4%9A%E5%88%86%E7%B1%BB%E9%97%AE%E9%A2%98convolution-neural-network/import torch from torchvision import transforms from torchvision import datasets from torch.utils.data import DataLoader import torch.nn.funhttps://blog.overfit.org/posts/2021-09-06-15235991-mnist%E6%89%8B%E5%86%99%E6%95%B0%E5%AD%97%E9%9B%86%E7%9A%84%E5%A4%9A%E5%88%86%E7%B1%BB%E9%97%AE%E9%A2%98linear-layer/Mon, 06 Sep 2021 22:10:00 +0800https://blog.overfit.org/posts/2021-09-06-15235991-mnist%E6%89%8B%E5%86%99%E6%95%B0%E5%AD%97%E9%9B%86%E7%9A%84%E5%A4%9A%E5%88%86%E7%B1%BB%E9%97%AE%E9%A2%98linear-layer/import torch # 引入模块PyTorch from torchvision import transforms # 从torch视觉中引入转换函数 from torchvision import datasets # 导入数据库 from torch.utils.dahttps://blog.overfit.org/posts/2021-09-06-15235875-pytorch%E5%88%9B%E5%BB%BA%E7%AE%80%E5%8D%95%E7%9A%84%E9%80%BB%E8%BE%91%E5%9B%9E%E5%BD%92%E6%A8%A1%E5%9E%8Blogisticregression/Mon, 06 Sep 2021 21:39:00 +0800https://blog.overfit.org/posts/2021-09-06-15235875-pytorch%E5%88%9B%E5%BB%BA%E7%AE%80%E5%8D%95%E7%9A%84%E9%80%BB%E8%BE%91%E5%9B%9E%E5%BD%92%E6%A8%A1%E5%9E%8Blogisticregression/import torch import torch.nn.functional as F # 从torch引入激活函数 x_data = torch.tensor([[1.0], [2.0], [3.0]]).cuda() # 将数据放在GPU上 y_data = torch.thttps://blog.overfit.org/posts/2021-09-06-15234803-pytorch%E5%AE%9E%E7%8E%B0linearmodel/Mon, 06 Sep 2021 17:34:00 +0800https://blog.overfit.org/posts/2021-09-06-15234803-pytorch%E5%AE%9E%E7%8E%B0linearmodel/这里选择将模型和loss还有数据都在GPU上，方便后续能够将大型神经网络的计算图放在GPU上来加快训练速度 import torch # 引入PyTorch模块 x = torch.tensor([[1.0], [2.0], [3.0]]).cuda() # 创建张量x，y并使用