yao62995/tensorflow
A collection of Chinese blog posts with diagrams explaining TensorFlow's internal source code structure.

This repository contains a series of educational articles analyzing TensorFlow’s source code through diagrams and technical explanations. The content covers TensorFlow’s core components including Session execution flow, Graph construction and optimization, device placement algorithms, memory allocation (BFC), thread pools, and kernel implementations for operations like MatMul and Conv2d. The material is presented as a study guide for developers seeking to understand TensorFlow internals.