Chainer
Chainer Tutorial
Introduction to Chainer
Core Concept
Forward/Backward Computation
Parameterized functions
FunctionSet
Optimizer
Example: Multi-layer Perceptron on MNIST
Recurrent Nets and their Computational Graph
Recurrent Nets
Truncate the Graph by Unchaining
Network Evaluation without Storing the Computation History
Using GPU(s) in Chainer
Relationship between Chainer and PyCUDA
Basics of GPUArray in Chainer
Run Neural Networks on a Single GPU
Model-parallel Computation on Multiple GPUs
Data-parallel Computation on Multiple GPUs
Define your own function
Non-parameterized Functions
Write an Elementwise Kernel Function
Parameterized Functions
Testing Function
Chainer Reference Manual
Core functionalities
Variable
Function
FunctionSet
Optimizer
Utilities
CUDA utilities
Initialization and global states
Devices and contexts
GPUArray allocation and copy
Random number generators
Kernel definition utilities
Interprocess communication on GPU
Common algorithms
Assertion and Testing
Type checking utilites
Gradient checking utilities
Standard Function implementations
Learnable connections
Array commputation functions
Array manipulation functions
Array computations
Activation functions
Pooling functions
Normalization functions
Loss, evaluation and aggregation
Reusable subnetwork of complex architectures
Optimizers
Caffe Reference Model Support
Visualization of Computational Graph
Chainer Contribution Guide
Classification of Contributions
Release and Milestone
Issues and PRs
Coding Guidelines
Testing Guidelines
Chainer
Docs
»
Chainer Reference Manual
»
Utilities
Edit on Bitbucket
Utilities
¶
CUDA utilities
¶
Initialization and global states
¶
Devices and contexts
¶
GPUArray allocation and copy
¶
Random number generators
¶
Kernel definition utilities
¶
Interprocess communication on GPU
¶
Common algorithms
¶