kanyun-inc/ytk-mp4j

Ytk-mp4j is a fast, user-friendly, cross-platform, multi-process, multi-thread collective message passing java library which includes gather, scatter, allgather, reduce-scatter, broadcast, reduce, allreduce communications for distributed machine learning.

45
/ 100
Emerging

This is a Java library designed for developers building distributed machine learning systems. It helps manage how data and messages are exchanged between many different computing processes and threads. Developers provide their raw data or complex Java objects, and the library handles the underlying communication to perform operations like aggregating or distributing information across the system.

111 stars. No commits in the last 6 months.

Use this if you are a Java developer building a distributed machine learning application and need to efficiently manage collective communication patterns across multiple processes and threads.

Not ideal if you are an end-user looking for a pre-built machine learning framework or if you are not working in a Java development environment.

distributed-computing machine-learning-engineering high-performance-computing java-development
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 9 / 25
Maturity 16 / 25
Community 20 / 25

How are scores calculated?

Stars

111

Forks

27

Language

Java

License

MIT

Last pushed

Jun 14, 2017

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/kanyun-inc/ytk-mp4j"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.