diffdist

condav0.1

Provides differentiable communication primitives such as Send/Recv/Gather etc for pytorch. Autograd can backpropagate through those, making distributed model parallel models easy to implement.

License GPL-3.0-or-later1 versions1 maintainers0 deps17 weekly dl
53
/ 100
Health
safe to use

[email protected] is safe to use (health: 53/100)

Health breakdown0 – 100
20/25
maintenance
0/20
popularity
25/25
security
6/15
maturity
2/15
community
Vulnerabilities
0
none known

Health History

Dependency Tree

License Audit

API access

Get this data programmatically — free, no authentication.

curl https://depscope.dev/api/check/conda/diffdist

First published · 2025-02-11 08:58:08.141000+00:00

Last updated · 2026-02-12 21:27:31.337000+00:00