Calcular的博客

本文部分参考 https://morvanzhou.github.io/https://www.bilibili.com/video/av9156347/?from=search&seid=14443754263386295565 如何安装?win10-x64下cpu版本1.下载安装python3.6勾选ADD环境变量。（此安装包自带numpy与pip）2.cm...

参考与评述参考书目《Deep Learning》Lan Goodfellow. 经典的深度学习框架是以计算图&梯度下降方法实现对前馈网络的有监督学习。 这里复现了前馈计算图的梯度计算实现。一、前馈计算图实现

一、尝试MNIST数据集的分类问题1. 导入MNIST数据集：#首先创建一个文件放在与主文件相同目录下命名为input_data.py内容如下"""Functions for downloading and reading MNIST data."""from __future__ import absolute_importfrom

tensor.h：/*This file is important that must be included*/#ifndef TENSOR_H#define TENSOR_H#define GPU_EN //Enable GPU or parallel-computing-device#define STATIC_TENSOR 0 //tensors have the...

train.py:# -*- coding: utf-8 -*-import torchimport torch.nn as nnimport cv2import numpy as npfrom torchvision import modelsimport torchvision.transforms as transformsfrom matplotlib import py...