Riemann Documentation

中文文档: https://riemann-zh.readthedocs.io/zh-cn/latest/

Riemann is an open-source lightweight automatic differentiation library designed specifically for learning, educational, and research purposes related to automatic differentiation and neural networks.

Getting Started

• Introduction
  • What is Riemann?
  • Key Features
  • Riemann Package Module Structure
  • Module Import Examples
  • Application Scenarios
  • PyTorch Compatibility
  • Installation Guide

Tensors

• Tensor Basics
  • What is a Tensor
  • Creating Tensors
  • Tensor Attributes and States
  • Tensor Computations
  • Tensor Type Conversion
  • Gradient Tracking
  • Indexing Operations
  • In-place Operations
  • Diagonalization Operations
  • Saving and Loading Tensors
  • Einstein Summation Convention (einsum)
• Einstein Summation Convention (einsum)
  • What is the Einstein Summation Convention
  • einsum Equation String Syntax
  • Classification of einsum Computation Scenarios
  • einsum Usage Examples
  • einsum Performance Optimization Tips
  • Notes

Automatic Differentiation

• Automatic Differentiation Basics
  • Gradient Tracking Switch
  • Computing Gradients
  • Gradient Accumulation
  • Gradient Computation Context Control
  • Tensor Methods for Graph Detaching and Data Copying
  • In-place Operations and Gradients
  • Higher-Order Gradients
  • Gradient functions (Functional API)
  • Custom Gradient Functions
  • Advanced Computational Graph Manipulation
  • Gradient Checking
  • Gradient Computation Tips
  • Common Pitfalls
  • Examples

Neural Networks

• How to Build a Neural Network
  • Step 1: Data Preparation
  • Step 2: Building the Neural Network
  • Step 3: Training the Network
  • Step 4: Evaluation and Inference
  • Step 5: Complete Example
  • Key Concepts Summary
• Module Class and Containers
  • Module Class Core Functionality
  • Module Class Main Methods
  • Creating Custom Modules
  • Container Classes
• Activation Functions
  • Activation Function List
• Loss Functions
  • Loss Function List
• Initialization Module
  • Initialization Functions List
  • Gain Value Reference Table
  • Usage Examples
• Basic Network Layers
  • Linear Layer (Linear)
  • Dropout Layer
  • Dropout2d Layer
  • Dropout3d Layer
  • Flatten Layer
  • BatchNorm1d Layer
  • BatchNorm2d Layer
  • BatchNorm3d Layer
  • LayerNorm Layer
  • Embedding Layer
• Module Hook Management
  • Hook Types Overview
  • Forward Pre-Hook (register_forward_pre_hook)
  • Forward Hook (register_forward_hook)
  • Full Backward Pre-Hook (register_full_backward_pre_hook)
  • Full Backward Hook (register_full_backward_hook)
  • Hook Registration and Removal
  • Typical Application Scenarios
  • Important Notes
• Convolutional Networks
  • Convolution Layers
  • Pooling Layers
  • MNIST Handwritten Digit Recognition Example
  • CIFAR-10 Image Classification Example
  • CNN Design Guidelines
• Transformer
  • Transformer Architecture Overview
  • MultiheadAttention Mechanism
  • Transformer Encoder
  • Transformer Decoder
  • Complete Transformer Model
  • Machine Translation Example
  • Differences Between Encoder and Decoder
  • Training and Inference Workflow
  • Transformer Design Guidelines
• KAN Networks
  • KAN Network Principles
  • KAN Network Application Scenarios
  • Riemann’s KAN Module
  • KANLinear Module
  • KAN Container Module
  • KAN Design Guidelines

Optimization

• Optimizers
  • Optimizer Basics
  • GD (Gradient Descent)
  • SGD (Stochastic Gradient Descent)
  • Adam (Adaptive Moment Estimation)
  • Adagrad
  • LBFGS
  • AdamW (Adam with Weight Decay)
  • RMSprop (Root Mean Square Propagation)
  • Optimizer Methods
  • Parameter Groups
  • Gradient Clipping
  • Examples
  • Custom Optimizer
• Learning Rate Schedulers
  • Types of Learning Rate Schedulers
  • Using Learning Rate Schedulers
  • Scheduler and Optimizer Interaction
  • Complete Example Code
  • Best Practices for Learning Rate Schedulers

Computer Vision

• Computer Vision
  • Overview
  • Quick Start
  • Datasets
  • Transforms
  • Complete Examples

Linear Algebra

• Linear Algebra
  • Matrix Operations
  • Decompositions
  • Norms and Vector Products
  • Linear Systems
  • Special Matrices
  • Examples

API Reference

• API Reference

Community & Contributing

• Contributing
• License

Indices and tables

• Index

• Module Index

• Search Page