Sparse-dense operators implementation for Paddle

Overview

Sparse-dense operators implementation for Paddle

This module implements coo, csc and csr matrix formats and their inter-ops with dense matrices.

Feel free to open an issue when you feel that something is incorrect.

Requirements

It only needs paddle. It is tested on paddle >= 2.1.0, <= 2.2.0rc1, but should work for any recent paddle versions.

Usage

Most functions are implemented within classes that encapsulate sparse formats: COO, CSR and CSC.

Cross-format operators are implemented in dedicated sub-modules: spgemm and batching.

Supported operations

Conversion

coo -> csc, csr, dense
csc -> coo
csr -> coo

Batch MVP (Matrix-Vector Product) or SpMM (Sparse-Dense Matmul)

Note that in this library, the batch dimensions are appended instead of prepended to the dot dimension (which makes batch MVP essentially regular matmul). Use utils.swap_axes or paddle.transpose when necessary.

coo, dense -> dense

Point-wise

Supports broadcast on the dense side.

coo + coo -> coo
coo * scalar -> coo
coo * dense -> coo (equiv. coo @ diag(vec) if dense is a vector)

SpGEMM (Sparse-Sparse Matmul)

coo, csr -> coo (via row-wise mixed product)

Batching and unbatching

Many batched operations can be efficiently represented via operation on block-diagonal sparse matrix. We also provide batching and unbatching operations for homogeneously-shaped sparse matrices.

For COO matrices, this is constructing (destructing) a block-diagonal COO matrix given (into) several small COO matrices.

If you know the expected shapes of matrices after unbatching you may construct it explicitly by calling BatchingInfo(shapes: [n, 2] numpy array of int). Otherwise: 1) most operations keep shapes, and there is no need to change BatchingInfo; 2) batch_info_dot is provided, for merging info between two batches of matrices that go through SpGeMM to obtain a final batch of matrices.

batch [coo] -> coo
unbatch coo -> [coo]

Installation

pip install paddle-sparse-dense

Caveats

Currently all stuff is implemented with pure python and no CUDA code has been written. As a result, the routines have good run-time performance in general but have a memory overhead of order O(nnz/n).

You might also like...
Paddle pit - Rethinking Spatial Dimensions of Vision Transformers
Paddle pit - Rethinking Spatial Dimensions of Vision Transformers

基于Paddle实现PiT ——Rethinking Spatial Dimensions of Vision Transformers,arxiv 官方原版代

Paddle-Skeleton-Based-Action-Recognition - DecoupleGCN-DropGraph, ASGCN, AGCN, STGCN

Paddle-Skeleton-Action-Recognition DecoupleGCN-DropGraph, ASGCN, AGCN, STGCN. Yo

Several simple examples for popular neural network toolkits calling custom CUDA operators.
Several simple examples for popular neural network toolkits calling custom CUDA operators.

Neural Network CUDA Example Several simple examples for neural network toolkits (PyTorch, TensorFlow, etc.) calling custom CUDA operators. We provide

Learning nonlinear operators via DeepONet

DeepONet: Learning nonlinear operators The source code for the paper Learning nonlinear operators via DeepONet based on the universal approximation th

Example repository for custom C++/CUDA operators for TorchScript

Custom TorchScript Operators Example This repository contains examples for writing, compiling and using custom TorchScript operators. See here for the

Jittor is a high-performance deep learning framework based on JIT compiling and meta-operators.
Jittor is a high-performance deep learning framework based on JIT compiling and meta-operators.

Jittor: a Just-in-time(JIT) deep learning framework Quickstart | Install | Tutorial | Chinese Jittor is a high-performance deep learning framework bas

Official implementation of our paper
Official implementation of our paper "LLA: Loss-aware Label Assignment for Dense Pedestrian Detection" in Pytorch.

LLA: Loss-aware Label Assignment for Dense Pedestrian Detection This project provides an implementation for "LLA: Loss-aware Label Assignment for Dens

Official implementation of Monocular Quasi-Dense 3D Object Tracking
Official implementation of Monocular Quasi-Dense 3D Object Tracking

Monocular Quasi-Dense 3D Object Tracking Monocular Quasi-Dense 3D Object Tracking (QD-3DT) is an online framework detects and tracks objects in 3D usi

Tensorflow implementation of the paper "HumanGPS: Geodesic PreServing Feature for Dense Human Correspondences", CVPR 2021.

HumanGPS: Geodesic PreServing Feature for Dense Human Correspondences Tensorflow implementation of the paper "HumanGPS: Geodesic PreServing Feature fo

Owner
北海若
Undergraduate, at SJTU & MSRA.
北海若
Differentiable Neural Computers, Sparse Access Memory and Sparse Differentiable Neural Computers, for Pytorch

Differentiable Neural Computers and family, for Pytorch Includes: Differentiable Neural Computers (DNC) Sparse Access Memory (SAM) Sparse Differentiab

ixaxaar 302 Dec 14, 2022
Paddle implementation for "Highly Efficient Knowledge Graph Embedding Learning with Closed-Form Orthogonal Procrustes Analysis" (NAACL 2021)

ProcrustEs-KGE Paddle implementation for Highly Efficient Knowledge Graph Embedding Learning with Orthogonal Procrustes Analysis ?? A more detailed re

Lincedo Lab 4 Jun 9, 2021
Paddle implementation for "Cross-Lingual Word Embedding Refinement by ℓ1 Norm Optimisation" (NAACL 2021)

L1-Refinement Paddle implementation for "Cross-Lingual Word Embedding Refinement by ℓ1 Norm Optimisation" (NAACL 2021) ?? A more detailed readme is co

Lincedo Lab 4 Jun 9, 2021
SplineConv implementation for Paddle.

SplineConv implementation for Paddle This module implements the SplineConv operators from Matthias Fey, Jan Eric Lenssen, Frank Weichert, Heinrich Mül

北海若 3 Dec 29, 2021
PaddleRobotics is an open-source algorithm library for robots based on Paddle, including open-source parts such as human-robot interaction, complex motion control, environment perception, SLAM positioning, and navigation.

简体中文 | English PaddleRobotics paddleRobotics是基于paddle的机器人开源算法库集,包括人机交互、复杂运动控制、环境感知、slam定位导航等开源算法部分。 人机交互 主动多模交互技术TFVT-HRI 主动多模交互技术是通过视觉、语音、触摸传感器等输入机器人

null 185 Dec 26, 2022
用强化学习DQN算法,训练AI模型来玩合成大西瓜游戏,提供Keras版本和PARL(paddle)版本

用强化学习玩合成大西瓜 代码地址:https://github.com/Sharpiless/play-daxigua-using-Reinforcement-Learning 用强化学习DQN算法,训练AI模型来玩合成大西瓜游戏,提供Keras版本、PARL(paddle)版本和pytorch版本

null 72 Dec 17, 2022
Paddle-Adversarial-Toolbox (PAT) is a Python library for Deep Learning Security based on PaddlePaddle.

Paddle-Adversarial-Toolbox Paddle-Adversarial-Toolbox (PAT) is a Python library for Deep Learning Security based on PaddlePaddle. Model Zoo Common FGS

AgentMaker 17 Nov 8, 2022
x-transformers-paddle 2.x version

x-transformers-paddle x-transformers-paddle 2.x version paddle 2.x版本 https://github.com/lucidrains/x-transformers 。 requirements paddlepaddle-gpu==2.2

yujun 7 Dec 8, 2022
This is the paddle code for SeBoW(Self-Born wiring for neural trees), a kind of neural tree born form a large search space

SeBoW: Self-Born Wiring for neural trees(PaddlePaddle version) This is the paddle code for SeBoW(Self-Born wiring for neural trees), a kind of neural

HollyLee 13 Dec 8, 2022
YOLOX-Paddle - A reproduction of YOLOX by PaddlePaddle

YOLOX-Paddle A reproduction of YOLOX by PaddlePaddle 数据集准备 下载COCO数据集,准备为如下路径 /ho

QuanHao Guo 6 Dec 18, 2022