The spelled-out intro to neural networks and backpropagation: building micrograd

推荐理由

It's especially useful because he explains the why and provides intuitive explanations instead of just talking about the what and how.

强烈推荐radq · 原文2023年6月

说明

这是2022年由Andrej Karpathy发布的一部课堂讲座，主题归在Machine learning & deep learning。它讲的是backpropagation and neural networks这一支里的具体问题，而不是把整个领域扫一遍。适合第一次接触、还没有先修课的人。内容上可以概括成：Andrej Karpathy's first Neural Networks: Zero to Hero lecture, which builds micrograd, a tiny autograd engine, from scratch in Python. It explains backpropagation and neural network training step by step, assuming only basic Python and some high-school calculus.这些读者给出的理由，核心是：HN treats it as the best starting point for really understanding backprop and neural nets, because Karpathy explains the why with intuition and builds everything in code.讲法不绑某一条新闻，过几年仍然可以当材料看。从2022年到2025年，有17位不同的读者在16个不同月份把它推荐出来，其中明确叫别人去看的有10次。节奏接近上课，可以暂停做笔记。它假设你愿意跟着一条推导走完。如果已经会了基本定义，就把注意力放在它怎么把backpropagation and neural networks变成能跟着复述的例子。

视频信息

频道

Andrej Karpathy

时长

2 小时 25 分

发布年份

2022

同主题更多

• But what is a neural network? | Deep learning chapter 1
• AlphaGo - The Movie | Full award-winning documentary
• Transformers, the tech behind LLMs | Deep Learning Chapter 5
• USENIX Security '18-Q: Why Do Keynote Speakers Keep Suggesting That Improving Security Is Possible?
• Lecture 1 | Machine Learning (Stanford)
• MarI/O - Machine Learning for Video Games