当前位置:网站首页>GNN third party Library: pyg (pytorch geometric) [the library based on pytorch can help users quickly build and train their own graph neural network model] [deepwalk, line, GCN, gat, etc.]
GNN third party Library: pyg (pytorch geometric) [the library based on pytorch can help users quickly build and train their own graph neural network model] [deepwalk, line, GCN, gat, etc.]
2022-07-23 10:09:00 【u013250861】
One . What is? PyG?
PyG The full name is Pytorch Geometric, It is based on Pytorch Constructed library , It can help users quickly build and train their own graph neural network model .
PyG Many advanced technologies have been achieved in (state of the art) Of GNN modular , for example GCN、GraphSage、GAT、SGC、GIN wait . Besides ,PyG It also contains a large number of benchmark Graph data set 、 Rich graph data operation methods 、 Support more GPU.
Two .PyG Installation
About PyG Installation , Recommended pip The way of command :
pip install torch-scatter -f https://data.pyg.org/whl/torch-${
TORCH}+${
CUDA边栏推荐
猜你喜欢

适合短视频分享的Fireshare

Baidu map eagle eye track service

spark分区算子partitionBy、coalesce、repartition

组合数。。。。

Multi UA V cooperative exploring for the unknown interior environment based on dynamic target tracking

【Azure 事件中心】Azure Event Hub 新功能尝试 -- 异地灾难恢复 (Geo-Disaster Recovery)

This tool complements the last kilometer of JMeter performance analysis

实现多层级条件查询(类似京东多层级添加查询)

Interviewer: explain the core principle of ThreadLocal

九张图纵观加密市场周期规律
随机推荐
This tool complements the last kilometer of JMeter performance analysis
华泰证劵开户有骗局吗安全吗
海通证券场内基金开户怎么样安全吗
目前都有哪些年利率6%左右的保本理财产品?
解密 Redis 助力双 11 背后电商秒杀系统
30行自己写并发工具类(Semaphore, CyclicBarrier, CountDownLatch)是什么体验?
BGP reflector, federal, routing rules
Technology sharing | big transaction blocking show master status
BGP反射器,联邦,选路规则
如何将list中相同字段值归类在同一个list下
Want to give up software testing, 4 years of experience to interview 10 minutes, the test is now so difficult?
Ten years of sharpening a sword, the core technology evolution of the cloud native distributed database polardb-x
STM32——输入捕获实验
C language flexible array
射频电路循证设计
指针的底层机制
华泰证券可以网上开户吗安全吗
《nlp入门+实战:第一章:深度学习和神经网络》
【机器学习基础】特征工程常用操作
网络通信原理与IP地址的分配原理,网络七层由下往上分别为物理层、数据链路层、网络层、传输层、会话层、表示层和应用层