本仓库旨在记录自己学习Pytorch的过程与收获同时帮助Pytorch初学者更快的入门。笔记分为三部分，分别为【学习笔记notebooks】、【参考模板reference】及【模型复现model】。
同步更新至博客ZRAINJ。
*全部笔记仅有中文版本，不定期更新。
其中记录了我根据教学视频初学Pytorch的笔记，记录了课堂重点内容并将代码重新敲了一遍。
已更新目录：
- 用Pytorch实现线性回归
- 用Pytorch加载数据集
- 用Pytorch实现多分类
总结了Pytorch各功能模板，方便自己构建新网络架构或迁移学习时参考快速便捷地实现Pytorch子功能模块。
已更新目录：
- Pytorch载入导出权重、冻结权重模板大全
仅利用少量库和接口，从零复现经典模型网络架构，对Pytoch有较高掌握要求。
已更新目录：
- YOLOv4
This repository aims to record the process and harvest of learning pytoch, and help pytoch beginners get started faster. The notes are divided into three parts, namely【notebooks】, 【reference】and【model】.
Sync update to blog ZRAINJ.
*All notes are only available in Chinese version, and are updated from time to time.
It recorded my notes on learning pytorch during the courses. I recorded the key contents of the class and gave the code.
contents:
- Linear regression with pytoch
- Dataset and dataloader
- Softmax classifier
The functional templates of pytoch are summarized to facilitate their reference when building a new network architecture or doing transfer learning, and quickly and conveniently realize the sub functional modules of pytoch.
contents:
- Complete set of Pytorch templates for exporting and importing weights and freezing weights
Only a few libraries are used to reproduce the classical model network architecture from scratch, which requires a high mastery of pytoch.
contents:
- YOLOv4