程序员老王
5.06万位订阅者
订阅者
30
音频
一个喜欢分享知识的普通程序员。 咨询服务与商务合作:[email protected] 个人主页:codewithwang.com.
点上方 「订阅 RSS」 自动复制订阅链接,粘贴到 Apple Podcasts、小宇宙、摸鱼 等任一播客 App 即可追更。 使用流程
音频 (30)

The Secret Behind 1M Context Windows: Sparse Attention, GQA, and KV Cache

Herdr: The Best Companion for VibeCoding

Pi: The Minimalist Coding Tool That’s Vibe-Checking Your Workflow

DLSS FSR 到底做了什么

What Exactly Is NVIDIA's Moat?

What is Vibe Coding? A Deep Dive into AI Programming Tools, from Models and Agents to Workflows

Python 3.15有什么新特性

How Large Models Understand Images: ViT (Vision Transformer)

This is how VibeCoding should be done!

AI的安全对齐,可能是个幻觉

What are prompt word engineering, context engineering, and Harness engineering?
![注意力残差是什么? [白话读论文]](/api/img?u=https%3A%2F%2Fi.ytimg.com%2Fvi%2FpGYrWsNQ8A0%2Fhqdefault.jpg)
注意力残差是什么? [白话读论文]

什么是LoRA 大模型微调是怎么回事

Deploy large models locally! Run DeepSeek-R1 with the Transformers library

Why is MoE evolving so rapidly? — The evolutionary history from elementary school math to the MoE...

What is MultiHeadAttention?

Understand LLM Skill in 10 minutes

Understand Tokens and Embeddings in 15 Minutes: A Detailed Explanation of LLM and RAG Data Proces...
![Training a Handwritten Digit Recognition Model with 30 Lines of Code [PyTorch in Action]](/api/img?u=https%3A%2F%2Fi.ytimg.com%2Fvi%2FqL6ca-mIeMI%2Fhqdefault.jpg)
Training a Handwritten Digit Recognition Model with 30 Lines of Code [PyTorch in Action]

Training principles of large models: Gradient descent: starting with a straight line

从零搭建神经网络,识别手写数字【PyTorch】【Transformer结构拆解】

从Linear到FeedForward AI模型的数学本质【Transformer结构拆解】

Transformer如何成为AI模型的地基

如何用腾讯EdgeOne 免费搭建个人博客

如何使用pytest进行单元测试
![AI思维链是幻象吗?[白话读论文]](/api/img?u=https%3A%2F%2Fi.ytimg.com%2Fvi%2FZLDfTwHm56A%2Fhqdefault.jpg)
AI思维链是幻象吗?[白话读论文]

如何使用Keynote快速制作动画?

专属研究员已就位,从查论文到做PPT,一人搞定全流程?!

一段提示词 让Gemini CLI变成自动化Agent! 提示词工程

AI 提示词工程 上下文工程 15分钟弄懂!