周健文's starred repositories
sealos
Sealos is a production-ready Kubernetes distribution that provides a one-stop solution for both public and private cloud. You can run any Docker image on sealos, start high availability databases like mysql/pgsql/redis/mongo, develop backend applications using node.js serverless
cpp-httplib
A C++ header-only HTTP/HTTPS server and client library
concurrentqueue
A fast multi-producer, multi-consumer lock-free concurrent queue for C++11
L-ink_Card
Smart NFC & ink-Display Card
tensorflow_template_application
TensorFlow template application for deep learning
LabelMeAnnotationTool
Source code for the LabelMe annotation tool.
gpushare-scheduler-extender
GPU Sharing Scheduler for Kubernetes Cluster
simple_tensorflow_serving
Generic and easy-to-use serving service for machine learning models
terracotta
A light-weight, versatile XYZ tile server, built with Flask and Rasterio :earth_africa:
WorldWideWeb
Last publicly available revision of the world's first web browser. This is a source import from 0.15 for NextStep. Originally written by @timbl.
WebRtc.NET
WebRTC for C# & C++/CLI
pyresample
Geospatial image resampling in Python
dask-image
Distributed image processing
apiAutoTest
Python+Requests+jsonpath+xlrd接口自动化测试工具,实现数据依赖,支持restful规范,sql断言以及测试前后数据隔离操作,自定义扩展方法,可作用于用例当中;video https://www.bilibili.com/video/BV1rr4y1r754/?vd_source=f824feef5305252d9a349520313fd4db
page-ruler-redux
An awesome page ruler extension for google chrome
tensorflow-serving-tutorial
A tutorial of building tensorflow serving service from scratch
grpc-nebula-c
微服务治理框架C++实现
cog-best-practices
Best practices with cloud-optimized-geotiffs (COGs)
consistent-hashing
Consistent Hashing implementation in Python
monitor-python
该项目为硬件实时监控系统,应用python、mysql、tornado、sqlalchemy、psutil、pyecharts等技术打造!
tf_serving_cpp_client
C++ client of a GAN model hosted by TensorFlow Serving
grpc-nebula-samples-c
微服务治理框架使用示例(C++版本)
image_cutting
Predicting DL model outputs for large images by cutting to smaller chunks
RasterIOCut
RasterIOCut
GraphN-GraphM-improved-version
随着现实世界中图处理需求的快速增长,大量迭代图处理作业同时在同一基础图上运行。而现有的并发图分析处理系统存在大量冗余数据存储和访问开销现有的并发图分析处理系统的存储系统则存在块表信息过大,对内存的利用效率不高等方面的问题,这些问题的原因一方面是块表信息的数据结构不合理,另一方面是块表信息过大后造成挤占可用内存空间的问题。 为解决此问题我们在现有的并发图分析处理系统的存储系统GraphM上实现了包括块表信息存储结构改进,块表优先级调度策略,轻量级的内存管理系统等多项重大改进。改进后的系统称作GraphN。实验表明:通过优化块表信息的数据结构,减少非必要的文件读写开销,将会造成更多读写开销的块缓存在内存中等方式,GraphN的块表信息仅占GraphM块表信息大小的千分之一,且同条件下同一任务所需时间缩短20%以上。