Skip to content

关于一些经典论文源码的逐行中文笔记

Notifications You must be signed in to change notification settings

RockyDing/Source-Code-Notebook

 
 

Repository files navigation

Source-Code-Notebook

关于一些经典论文源码Source Code的中文笔记,尽量做到整体框架梳理逐行源码注释

  • 不过太早期的代码一般都模块化了,所以主要更一些后期(2017-NOW)的笔记。

  • 语言主要是Pytorch和Tensorflow版本的代码,部分Keras。

  • 部分代码Run起来需要涉及到配置、数据集和预训练模型,请对应原文章的源码进行操作。

  • 有注解错误和思路问题欢迎讨论。

欢迎访问博客:https://nakaizura.blog.csdn.net/

推荐系统

Recommendation Notebook keywords
BPR Notebook Pair-wise
NFM Notebook MLP+FM
AFM Notebook Attention+FM
NCF Notebook MLP+MF
ONCF Notebook Out-product+NCF
DIN Notebook Attention+Interest
DRL-REC Notebook RL in Rec
IRGAN Notebook GAN in Rec
RippleNet Notebook KG in Rec
NGCF Notebook Graph in Rec
LightGCN Notebook Light Vesion NGCF
SR-GNN Notebook Session-based GNN

图神经网络

Graph Notebook keywords
Louvain Notebook Community Detection
DeepWalk Notebook Graph Embedding
node2vec Notebook Graph Embedding Plus
GCN Notebook Spectral Convolutional
GAT Notebook Spatial+Attention
GraphSAGE Notebook Inductive
GAE Notebook Graph Autoencoder
HetGNN Notebook Heterogeneous
Scenegraph Notebook Relationship Triplet
Graph Transformer Notebook Graph Transformer

计算机视觉

CV Notebook keywords
DCGAN Notebook GAN in CV
InfoGAN Notebook Info Control
Capsules Notebook CNN is Rubbish
C3D Notebook 3D VGG
SlowFast Notebook Slow+Fast
Non-local Notebook Pixel+Attention
MoCo Notebook Contrastive Learning
MIL-NCE Notebook Contrastive Learning

自然语言处理

NLP Notebook keywords
NTM Notebook NVDM-GSM
Skip-Thought Notebook Sentence2vec
Transformer Notebook Attention is All
BERT Notebook Pretraning NLP

跨模态

Cross-modal Notebook keywords
TALL Notebook Cross-modal Video Moment Retrieval
ACMR Notebook Adversarial Cross-modal Retrieval
S2VT Notebook Cross-modal Video Captioning
VMT Notebook Cross-modal Video Translation
MIL-NCE Notebook Cross-modal Align
Preparing Notebook 待更......

About

关于一些经典论文源码的逐行中文笔记

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%